Patent Grant
10664287
The field of invention relates generally to computer processor architecture, and, more specifically, to systems and methods for implementing chained tile operations.
Matrices are increasingly important in many computing tasks such as machine learning and other bulk data processing.
The present invention is illustrated by way of example and not limitation in the figures of the accompanying drawings, in which like references indicate similar elements and in which:
In the following description, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In other instances, well-known circuits, structures and techniques have not been shown in detail in order not to obscure the understanding of this description.
References in the specification to “one embodiment,” “an embodiment,” “an example embodiment,” etc., indicate that the embodiment described may include a particular feature, structure, or characteristic, but every embodiment may not necessarily include the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Further, when a particular feature, structure, or characteristic is described in connection with an embodiment, it is submitted that it is within the knowledge of one skilled in the art to affect such feature, structure, or characteristic in connection with other embodiments whether or not explicitly described.
In many mainstream processors, handling matrices is a difficult and/or instruction intensive task. For example, rows of a matrix could be put into a plurality of packed data (e.g., SIMD or vector) registers and then operated on individually. For example, an add two 8×2 matrices may require a load or gather into four packed data registers depending upon data sizes. Then a first add of packed data registers corresponding to a first row from each matrix is performed and a second add of packed data registers corresponding to a second row from each matrix is performed. Then the resulting packed data registers are scattered back to memory. While for small matrices this scenario may be acceptable, it is often not acceptable with larger matrices.
I. High-Level Discussion
Described herein are mechanisms to support matrix operations in computer hardware such as central processing units (CPUs), graphic processing units (GPUs), and accelerators. The matrix operations utilize 2-dimensional (2-D) data structures representing one or more packed regions of memory such as registers. Throughout this description, these 2-D data structures are referred to as tiles. Note that a matrix may be smaller than a tile (use less than all of a tile) or utilize a plurality of tiles (the matrix is larger than the size of any one tile). Throughout the description, tile language is used to indicate operations performed using tiles that impact a matrix; whether or not that matrix is larger than any one tile is not typically relevant.
Each tile may be acted upon by different operations such as those that are detailed herein and include, but are not limited to: tile multiplication, tile add, tile subtract, tile diagonal, tile zero, tile transpose, tile dot product, tile broadcast, tile row broadcast, tile column broadcast, tile multiplication, tile multiplication and accumulation, tile move, etc. Additionally, support for operators such as the use of a scale and/or bias may be used with these operations or in support of non-numeric applications in the future, for instance, OpenCL “local memory,” data compression/decompression, etc.
Portions of storage (such as memory (non-volatile and volatile), registers, cache, etc.) are arranged into tiles of different horizontal and vertical dimensions. For example, a tile may have horizontal dimension of 4 (e.g., four rows of a matrix) and a vertical dimension of 8 (e.g., 8 columns of the matrix). Typically, the horizontal dimension is related to element sizes (e.g., 2-, 4-, 8-, 16-, 32-, 64-, 128-bit, etc.). Multiple datatypes (single precision floating point, double precision floating point, integer, etc.) may be supported.
A. Exemplary Usage of Configured Tiles
In some embodiments, tile parameters can be configured. For example, a given tile may be configured to provide tile options. Exemplary tile options include but are not limited to: a number of rows of the tile, a number of columns of the tile, whether the tile is VALID, and whether the tile consists of a PAIR of equal-sized tiles.
In some embodiments, tile parameters are definable. For example, a “palette” is used to provide tile options. Exemplary options include, but are not limited to: the number of tile names, the number of bytes in a row of storage, the number of rows and columns in a tile, etc. For example, a maximum “height” (number of rows) of a tile may be defined as:
Tile Max Rows=Architected Storage/(The Number of Palette Names*The Number of Bytes per row).
As such, an application can be written such that a fixed usage of names will be able to take advantage of different storage sizes across implementations.
Configuration of tiles is done using a tile configuration (“TILECONFIG”) instruction, where a particular tile usage is defined in a selected palette. This declaration includes the number of tile names to be used, the requested number of rows and columns per name (tile), and, in some embodiments, the requested datatype of each tile. In some embodiments, consistency checks are performed during the execution of a TILECONFIG instruction to determine that it matches the restrictions of the palette entry.
B. Exemplary Tile Storage Types
Tile loads from memory and stores to memory are typically strided accesses from the application memory to packed rows of data. Exemplary TILELOAD and TILESTORE instructions, or other instruction references to application memory as a TILE operand in load-op instructions, are, in some embodiments, restartable to handle (up to) 2*rows of page faults, unmasked floating point exceptions, and/or interrupts per instruction.
In (B), a matrix is stored in a tile comprised of a plurality of registers such as packed data registers (single instruction, multiple data (SIMD) or vector registers). In this example, the tile is overlaid on three physical registers. Typically, consecutive registers are used, however, this need not be the case.
In (C), a matrix is stored in a tile in non-register storage accessible to a fused multiple accumulate (FMA) circuit used in tile instructions. This storage may be inside of a FMA, or adjacent to it. Additionally, in some embodiments, discussed below, the storage may be for a data element and not an entire row or tile.
The supported parameters for the TMMA architecture are reported via CPUID. In some embodiments, the list of information includes a maximum height and a maximum SIMD dimension. Configuring the TMMA architecture requires specifying the dimensions for each tile, the element size for each tile and the palette identifier. This configuration is done by executing the TILECONFIG instruction.
Successful execution of a TILECONFIG instruction enables subsequent TILE operators. A TILERELEASEALL instruction clears the tile configuration and disables the TILE operations (until the next TILECONFIG instructions executes). In some embodiments, XSAVE, XSTORE, etc. are used in context switching using tiles. In some embodiments, 2 XCRO bits are used in XSAVE, one for TILECONFIF metadata and one bit corresponding to actual tile payload data.
TILECONFIG not only configures the tile usage, but also sets a state variable indicating that the program is in a region of code with tiles configured. An implementation may enumerate restrictions on other instructions that can be used with a tile region such as no usage of an existing register set, etc.
Exiting a tile region is typically done with the TILERELEASEALL instruction. It takes no parameters and swiftly invalidates all tiles (indicating that the data no longer needs any saving or restoring) and clears the internal state corresponding to being in a tile region.
In some embodiments, tile operations will zero any rows and any columns beyond the dimensions specified by the tile configuration. For example, tile operations will zero the data beyond the configured number of columns (factoring in the size of the elements) as each row is written. For example, with 64 byte rows and a tile configured with 10 rows and 12 columns, an operation writing FP32 elements would write each of the first 10 rows with 12*4 bytes with output/result data and zero the remaining 4*4 bytes in each row. Tile operations also fully zero any rows after the first 10 configured rows. When using 1K tile with 64 byte rows, there would be 16 rows, so in this example, the last 6 rows would also be zeroed.
In some embodiments, a context restore (e.g., XRSTOR), when loading data, enforces that the data beyond the configured rows for a tile will be maintained as zero. If there is no valid configuration, all rows are zeroed. XRSTOR of tile data can load garbage in the columns beyond those configured. It should not be possible for XRSTOR to clear beyond the number of columns configured because there is not an element width associated with the tile configuration.
Context save (e.g., XSAVE) exposes the entire TILE storage area when writing it to memory. If XRSTOR loaded garbage data in to the rightmost part of a tile, that data will be saved by XSAVE. XSAVE will write zeros for rows beyond the number specified for each tile.
In some embodiments, tile instructions are restartable. The operations that access memory allow restart after page faults. The computational instructions that deal with floating point operations also allow for unmasked floating point exceptions, with the masking of the exceptions controlled by a control and/or status register.
To support restarting instructions after these events, the instructions store information in the start registers detailed below.
II. Matrix (Tile) Operation Systems
A. Exemplary Hardware Support
In this example, a coherent memory interface 303 is coupled to the host processor/processing system 301 and matrix operations accelerator 307 such that they can share memory.
In some embodiments, the matrix operations accelerator 307 includes a plurality of FMAs 309 coupled to data buffers 305 (in some implementations, one or more of these data buffers 305 are stored in the FMAs of the grid as shown). The data buffers 305 buffer tiles loaded from memory and/or tiles to be stored to memory (e.g., using a tileload or tilestore instruction). Data buffers may be, for example, a plurality of registers. Typically, these FMAs are arranged as a grid of chained FMAs 309 which are able to read and write tiles. In this example, the matrix operations accelerator 307 is to perform a matrix multiply operation using tiles T0, T1, and T2. At least one of tiles is housed in the FMA grid 309. In some embodiments, all tiles in an operation are stored in the FMA grid 309. In other embodiments, only a subset is stored in the FMA grid 309. As shown, T1 is housed and T0 and T2 are not. Note that A, B, and C refer to the matrices of these tiles which may or may not take up the entire space of the tile.
The number of rows in the matrix (TILE A 601) matches the number of serial (chained) FMAs comprising the computation's latency. An implementation is free to recirculate on a grid of smaller height, but the computation remains the same.
The source/destination vector comes from a tile of N rows (TILE C 605) and the grid of FMAs 611 performs N vector-matrix operations resulting in a complete instruction performing a matrix multiplication of tiles. Tile B 603 is the other vector source and supplies “broadcast” terms to the FMAs in each stage.
In operation, in some embodiments, the elements of matrix B (stored in a tile B 603) are spread across the rectangular grid of FMAs. Matrix B (stored in tile A 601) has its elements of a row transposed to match up with the columnar dimension of the rectangular grid of FMAs. At each FMA in the grid, an element of A and B are multiplied and added to the incoming summand (from above in the Figure) and the outgoing sum is passed to the next row of FMAs (or the final output).
The latency of a single step is proportional to K (row height of matrix B) and dependent TMMAs typically have enough source-destination rows (either in a single tile or across tile) to hide that latency. An implementation may also split the SIMD (packed data element) dimension M (row height of matrix A) across time steps, but this simply changes the constant that K is multiplied by. When a program specifies a smaller K than the maximum enumerated by the TMACC, an implementation is free to implement this with “masking” or “early outs.”
The latency of an entire TMMA is proportional to N*K. The repeat rate is proportional to N. The number of MACs per TMMA instruction is N*K*M.
A first signed source (source 1701) and a second signed source (source 2703) each have four packed data elements. Each of these packed data elements stores signed data such as floating point data. A third signed source (source 3709) has two packed data elements, each of which stores signed data. The sizes of the first and second signed sources 701 and 703 are half that of the third signed source (initial value or previous result) 709. For example, the first and second signed sources 701 and 703 could have 32-bit packed data elements (e.g., single precision floating point) while the third signed source 709 could have 64-bit packed data elements (e.g., double precision floating point).
In this illustration, only the two most significant packed data element positions of the first and second signed sources 701 and 703 and the most significant packed data element position of the third signed source 709 are shown. Of course, the other packed data element positions would also be processed.
As illustrated, packed data elements are processed in pairs. For example, the data of the most significant packed data element positions of the first and second signed sources 701 and 703 are multiplied using a multiplier circuit 705, and the data from second most significant packed data element positions of the first and second signed sources 701 and 703 are multiplied using a multiplier circuit 707. In some embodiments, these multiplier circuits 705 and 707 are reused for other packed data elements positions. In other embodiments, additional multiplier circuits are used so that the packed data elements are processed in parallel. In some contexts, parallel execution is done using lanes that are the size of the signed third source 709. The results of each of the multiplications are added using adder 711 circuitry.
The result of the addition of the results of the multiplications is added to the data from most significant packed data element position of the signed source 3709 (using a different adder 713 or the same adder 711).
Finally, the result of the second addition is either stored into the signed destination 715 in a packed data element position that corresponds to the packed data element position used from the signed third source 709 or passed on to the next iteration if there is one. In some embodiments, a writemask is applied to this storage such that if a corresponding writemask (bit) is set, the storage happens, and, if not set, the storage does not happen.
A first signed source (source 1801) and a second signed source (source 2803) each have four packed data elements. Each of these packed data elements stores signed data such as integer data. A third signed source (source 3809) has two packed data elements, each of which stores signed data. The sizes of the first and second signed sources 801 and 803 are half that of the third signed source 809. For example, the first and second signed sources 801 and 803 could have 32-bit packed data elements (e.g., single precision floating point) the third signed source 809 could have 64-bit packed data elements (e.g., double precision floating point).
In this illustration, only the two most significant packed data element positions of the first and second signed sources 801 and 803 and the most significant packed data element position of the third signed source 809 are shown. Of course, the other packed data element positions would also be processed.
As illustrated, packed data elements are processed in pairs. For example, the data of the most significant packed data element positions of the first and second signed sources 801 and 803 are multiplied using a multiplier circuit 805, and the data from second most significant packed data element positions of the first and second signed sources 801 and 803 are multiplied using a multiplier circuit 807. In some embodiments, these multiplier circuits 805 and 807 are reused for other packed data elements positions. In other embodiments, additional multiplier circuits are used so that the packed data elements are processed in parallel. In some contexts, parallel execution is done using lanes that are the size of the signed third source (initial value or previous iteration result) 809. The results of each of the multiplications are added to the signed third source 809 using addition/saturation 813 circuitry.
Addition/saturation (accumulator) circuitry 813 preserves a sign of an operand when the addition results in a value that is too big. In particular, saturation evaluation occurs on the infinite precision result between the multi-way-add and the write to the destination or next iteration. When the accumulator 813 is floating point and the input terms are integer, the sum of products and the floating point accumulator input value are turned into infinite precision values (fixed point numbers of hundreds of bits), the addition of the multiplication results and the third input is performed, and a single rounding to the actual accumulator type is performed.
Unsigned saturation means the output values are limited to a maximum unsigned number for that element width (all 1s). Signed saturation means a value is limited to the be in the range between a minimum negative number and a max positive number for that element width (for bytes for example, the range is from −128 (=−2{circumflex over ( )}7) to 127(=2{circumflex over ( )}7−1)).
The result of the addition and saturation check is stored into the signed result 815 in a packed data element position that corresponds to the packed data element position used from the signed third source 809 or passed on to the next iteration if there is one. In some embodiments, a writemask is applied to this storage such that if a corresponding writemask (bit) is set, the storage happens, and, if not set, the storage does not happen.
A first signed source (source 1901) and a second unsigned source (source 2903) each have four packed data elements. Each of these packed data elements has data such as floating point or integer data. A third signed source (initial value or result 915) has a packed data element of which stores signed data. The sizes of the first and second sources 901 and 903 are a quarter of the third signed source 915. For example, the first and second sources 901 and 903 could have 16-bit packed data elements (e.g., word) and the third signed source 915 could have 64-bit packed data elements (e.g., double precision floating point or 64-bit integer).
In this illustration, the four most significant packed data element positions of the first and second sources 901 and 903 and the most significant packed data element position of the third signed source 915 are shown. Of course, other packed data element positions would also be processed if there are any.
As illustrated, packed data elements are processed in quadruplets. For example, the data of the most significant packed data element positions of the first and second sources 901 and 903 are multiplied using a multiplier circuit 905, data from second most significant packed data element positions of the first and second sources 901 and 903 are multiplied using a multiplier circuit 907, data from third most significant packed data element positions of the first and second sources 901 and 903 are multiplied using a multiplier circuit 909, and data from the least significant packed data element positions of the first and second sources 901 and 903 are multiplied using a multiplier circuit 911. In some embodiments, the signed packed data elements of the first source 901 are sign extended and the unsigned packed data elements of the second source 903 are zero extended prior to the multiplications.
In some embodiments, these multiplier circuits 905-911 are reused for other packed data elements positions. In other embodiments, additional multiplier circuits are used so that the packed data elements are processed in parallel. In some contexts, parallel execution is done using lanes that are the size of the signed third source 915. The results of each of the multiplications are added using addition circuitry 911.
The result of the addition of the results of the multiplications is added to the data from most significant packed data element position of the signed source 3915 (using a different adder 913 or the same adder 911).
Finally, the result 919 of the second addition is either stored into the signed destination in a packed data element position that corresponds to the packed data element position used from the signed third source 915 or passed to the next iteration. In some embodiments, a writemask is applied to this storage such that if a corresponding writemask (bit) is set, the storage happens, and, if not set, the storage does not happen.
A first signed source (signed source 11001) and a second unsigned source (unsigned source 21003) each have four packed data elements. Each of these packed data elements stores data such as floating point or integer data. A third signed source (initial or previous result 1015) has a packed data element of which stores signed data. The sizes of signed source 11001 and unsigned sourced 21003 are a quarter of the third signed source connected to initial or previous result 1015. For example, signed source 11001 and unsigned source 21003 could have 16-bit packed data elements (e.g., word) and initial or previous result 1015 could have 64-bit packed data elements (e.g., double precision floating point or 64-bit integer).
In this illustration, the four most significant packed data element positions of signed source 11001 and unsigned sourced 21003 and the most significant packed data element position of initial or previous result 1015 are shown. Of course, other packed data element positions would also be processed if there are any.
As illustrated, packed data elements are processed in quadruplets. For example, the data of the most significant packed data element positions of signed source 11001 and unsigned sourced 21003 are multiplied using a multiplier circuit 1005, data from second most significant packed data element positions of signed source 11001 and unsigned sourced 21003 are multiplied using a multiplier circuit 1007, data from third most significant packed data element positions of signed source 11001 and unsigned sourced 21003 are multiplied using a multiplier circuit 1009, and data from the least significant packed data element positions of signed source 11001 and unsigned sourced 21003 are multiplied using a multiplier circuit 1011. In some embodiments, the signed packed data elements of the signed source 11001 are sign extended and the unsigned packed data elements of unsigned source 21003 are zero extended prior to the multiplications.
In some embodiments, these multiplier circuits 1005-1011 are reused for other packed data elements positions. In other embodiments, additional multiplier circuits are used so that the packed data elements are processed in parallel. In some contexts, parallel execution is done using lanes that are the size of initial or previous result 1015. The result of the addition of the results of the multiplications is added to the data from most significant packed data element position of the signed source 3, connected to initial or previous result 1015 using addition/saturation circuitry 1013.
Addition/saturation circuitry 1013 preserves a sign of an operand when the addition results in a value that is too big or too small for signed saturation. As used herein, addition/saturation circuitry 1013 is sometimes referred to as an accumulator. In particular, saturation evaluation occurs on the infinite precision result between the multi-way-add and the write to the destination. When the addition/saturation circuitry 1013 uses floating point and the input terms are integer, the sum of products and the floating point accumulator input value are turned into infinite precision values (fixed point numbers of hundreds of bits), the addition of the multiplication results and the third input is performed, and a single rounding to the actual accumulator type is performed.
The result 1019 of the addition and saturation check is stored into the signed destination in a packed data element position that corresponds to the packed data element position used from the signed third source, connected to initial or previous result 1015, or passed to the next iteration. In some embodiments, a writemask is applied to this storage such that if a corresponding writemask (bit) is set, the storage happens, and, if not set, the storage does not happen.
For an accumulator having 4× input sizes (in other words, the accumulator input value is four times the size of the packed data element sizes of the sources), table 1103 illustrates different configurations. For byte sized sources, the accumulator uses 32-bit integer or single-precision floating-point (SPFP) values that are 32-bit in size. For word sized sources, the accumulator uses 64-bit integer or double-precision floating-point (DPFP) values that are 64-bit in size in some embodiments.
For an accumulator having 8× input sizes (in other words, the accumulator input value is eight times the size of the packed data element sizes of the sources), table 1105 illustrates a configuration. For byte sized sources, the accumulator uses 64-bit integer.
As hinted at earlier, matrix operations circuitry may be included in a core, or as an external accelerator.
A plurality of cores 1201, 1203, 1205, and 1207 provide non-tile based instruction support. In some embodiments, matrix operations circuitry 1251 is provided in a core 1203, and in other embodiments matrix operations circuitry 1211 and 1213 are accessible on the ring interconnect 1245.
Additionally, one or more memory controllers 1223-1225 are provided to communicate with memory 1233 and 1231 on behalf of the cores and/or matrix operations circuitry.
The branch prediction and decode circuitry 1303 is coupled to a rename/allocator circuitry 1307 which is coupled, in some embodiments, to scheduler circuitry 1309. In some embodiments, these circuits provide register renaming, register allocation, and/or scheduling functionality by performing one or more of: 1) renaming logical operand values to physical operand values (e.g., a register alias table in some embodiments), 2) allocating status bits and flags to the decoded instruction, and 3) scheduling the decoded instruction for execution on execution circuitry out of an instruction pool (e.g., using a reservation station in some embodiments).
The scheduler circuitry 1309 represents any number of different schedulers, including reservations stations, central instruction window, etc. The scheduler unit(s) scheduler circuitry 1309 is coupled to, or includes, physical register file(s) 1315. Each of the physical register file(s) 1315 represents one or more physical register files, different ones of which store one or more different data types, such as scalar integer, scalar floating point, packed integer, packed floating point, vector integer, vector floating point, status (e.g., an instruction pointer that is the address of the next instruction to be executed), tiles, etc. In one embodiment, the physical register file(s) 1315 comprises vector registers circuitry, write mask registers circuitry, and scalar registers circuitry. These register circuits may provide architectural vector registers, vector mask registers, and general purpose registers. The physical register file(s) 1315 is overlapped by a retirement circuit 1317 to illustrate various ways in which register renaming and out-of-order execution may be implemented (e.g., using a reorder buffer(s) and a retirement register file(s); using a future file(s), a history buffer(s), and a retirement register file(s); using a register maps and a pool of registers; etc.). The retirement circuit 1317 and the physical register file(s) 1315 are coupled to the execution circuit(s) 1311.
While register renaming is described in the context of out-of-order execution, it should be understood that register renaming may be used in an in-order architecture. While the illustrated embodiment of the processor may also include separate instruction and data cache units and a shared L2 cache unit, alternative embodiments may have a single internal cache for both instructions and data, such as, for example, a Level 1 (L1) internal cache, or multiple levels of internal cache. In some embodiments, the system may include a combination of an internal cache and an external cache that is external to the core and/or the processor. Alternatively, all of the cache may be external to the core and/or the processor.
The execution circuitry 1311 includes a set of one or more execution circuits 1321, 1323, and 1327 and a set of one or more memory access circuits 1325. The execution circuits 1321, 1323, and 1327 perform various operations (e.g., shifts, addition, subtraction, multiplication) and on various types of data (e.g., scalar floating point, packed integer, packed floating point, vector integer, vector floating point). While some embodiments may include a number of execution units dedicated to specific functions or sets of functions, other embodiments may include only one execution unit or multiple execution units that all perform all functions. The scalar circuitry 1321 performs scalar operations, the vector/SIMD circuitry 1323 performs vector/SIMD operations, and matrix operations circuitry 1327 performs matrix (tile) operations detailed herein.
By way of example, the exemplary register renaming, out-of-order issue/execution core architecture may implement a pipeline as follows: 1) an instruction fetch circuit performs fetch and length decoding stages; 2) the branch and decode circuitry 1303 performs a decode stage; 3) the rename/allocator circuitry 1307 performs an allocation stage and renaming stage; 4) the scheduler circuitry 1309 performs a schedule stage; 5) physical register file(s) (coupled to, or included in, the scheduler circuitry 1309 and rename/allocate circuitry 1307 and a memory unit perform a register read/memory read stage; the execution circuitry 1311 performs an execute stage; 6) a memory unit and the physical register file(s) unit(s) perform a write back/memory write stage; 7) various units may be involved in the exception handling stage; and 8) a retirement unit and the physical register file(s) unit(s) perform a commit stage.
The core may support one or more instructions sets (e.g., the x86 instruction set (with some extensions that have been added with newer versions); the MIPS instruction set of MIPS Technologies of Sunnyvale, Calif.; the ARM instruction set (with optional additional extensions such as NEON) of ARM Holdings of Sunnyvale, Calif.), including the instruction(s) described herein. In one embodiment, the core 1390 includes logic to support a packed data instruction set extension (e.g., AVX1, AVX2), thereby allowing the operations used by many multimedia applications to be performed using packed data.
It should be understood that the core may support multithreading (executing two or more parallel sets of operations or threads), and may do so in a variety of ways including time sliced multithreading, simultaneous multithreading (where a single physical core provides a logical core for each of the threads that physical core is simultaneously multithreading), or a combination thereof (e.g., time sliced fetching and decoding and simultaneous multithreading thereafter such as in the Intel® Hyperthreading technology).
The branch prediction and decode circuitry 1403 is coupled to a rename/allocator circuitry 1407 which is coupled, in some embodiments, to scheduler circuitry 1409. In some embodiments, these circuits provide register renaming, register allocation, and/or scheduling functionality by performing one or more of: 1) renaming logical operand values to physical operand values (e.g., a register alias table in some embodiments), 2) allocating status bits and flags to the decoded instruction, and 3) scheduling the decoded instruction for execution on execution circuitry out of an instruction pool (e.g., using a reservation station in some embodiments).
The scheduler circuitry 1409 represents any number of different schedulers, including reservations stations, central instruction window, etc. The scheduler unit(s) scheduler circuitry 1409 is coupled to, or includes, physical register file(s) 1415. Each of the physical register file(s) 1415 represents one or more physical register files, different ones of which store one or more different data types, such as scalar integer, scalar floating point, packed integer, packed floating point, vector integer, vector floating point, status (e.g., an instruction pointer that is the address of the next instruction to be executed), tiles, etc. In one embodiment, the physical register file(s) 1415 comprises vector registers circuitry, write mask registers circuitry, and scalar registers circuitry. These register circuits may provide architectural vector registers, vector mask registers, and general purpose registers. The physical register file(s) 1415 is overlapped by a retirement circuit 1417 to illustrate various ways in which register renaming and out-of-order execution may be implemented (e.g., using a reorder buffer(s) and a retirement register file(s); using a future file(s), a history buffer(s), and a retirement register file(s); using a register maps and a pool of registers; etc.). The retirement circuit 1417 and the physical register file(s) 1415 are coupled to execution circuitry 1411.
While register renaming is described in the context of out-of-order execution, it should be understood that register renaming may be used in an in-order architecture. While the illustrated embodiment of the processor may also include separate instruction and data cache units and a shared L2 cache unit, alternative embodiments may have a single internal cache for both instructions and data, such as, for example, a Level 1 (L1) internal cache, or multiple levels of internal cache. In some embodiments, the system may include a combination of an internal cache and an external cache that is external to the core and/or the processor. Alternatively, all of the cache may be external to the core and/or the processor.
The execution circuitry 1411 includes a set of one or more execution circuits 1427 and a set of one or more memory access circuits 1425. The execution circuits 1427 perform matrix (tile) operations detailed herein.
By way of example, the exemplary register renaming, out-of-order issue/execution core architecture may implement a pipeline as follows: 1) an instruction fetch circuit performs fetch and length decoding stages; 2) the branch and decode circuitry 1403 performs a decode stage; 3) the rename/allocator circuitry 1407 performs an allocation stage and renaming stage; 4) the scheduler circuitry 1409 performs a schedule stage; 5) physical register file(s) (coupled to, or included in, the scheduler circuitry 1409 and rename/allocate circuitry 1407 and a memory unit perform a register read/memory read stage; the execution circuitry 1411 performs an execute stage; 6) a memory unit and the physical register file(s) unit(s) perform a write back/memory write stage; 7) various units may be involved in the exception handling stage; and 8) a retirement unit and the physical register file(s) unit(s) perform a commit stage.
The core may support one or more instructions sets (e.g., the x86 instruction set (with some extensions that have been added with newer versions); the MIPS instruction set of MIPS Technologies of Sunnyvale, Calif.; the ARM instruction set (with optional additional extensions such as NEON) of ARM Holdings of Sunnyvale, Calif.), including the instruction(s) described herein. In one embodiment, the core 1490 includes logic to support a packed data instruction set extension (e.g., AVX1, AVX2), thereby allowing the operations used by many multimedia applications to be performed using packed data.
It should be understood that the core may support multithreading (executing two or more parallel sets of operations or threads), and may do so in a variety of ways including time sliced multithreading, simultaneous multithreading (where a single physical core provides a logical core for each of the threads that physical core is simultaneously multithreading), or a combination thereof (e.g., time sliced fetching and decoding and simultaneous multithreading thereafter such as in the Intel® Hyperthreading technology).
B. Layout
Throughout this description, data is expressed using row major data layout. Column major users should translate the terms according to their orientation.
In some embodiments, row-major semantics are utilized in hardware, and column major data is to swap the operand order with the result being transposes of matrix, but for subsequent column-major reads from memory it is the correct, non-transposed matrix.
For example, if there are two column-major matrices to multiply:
The input matrices would be stored in linear memory (column-major) as:
a c e b d f
and
g h i j k l.
Reading those matrices as row-major with dimensions 2×3 and 3×2, they would appear as:
Swapping the order and matrix multiplying:
the transpose matrix is out and can then be stored in in row-major order:
III. Exemplary Usage
The exemplary code as shown includes the usage of a tile configuration instruction and is executed to configure tile usage, load tiles, a loop to process the tiles, store tiles to memory, and release tile usage.
IV. Exemplary Configuration
A. Tile Configuration Hardware Support
As discussed above, tile usage typically needs to be configured prior to use. For example, full usage of all rows and columns may not be needed. Not only does not configuring these rows and columns save power in some embodiments, but the configuration may be used to determine if an operation will generate an error. For example, a matrix multiplication of the form (N×M)*(L*N) will typically not work if M and L are not the same.
Prior to using matrices using tiles, in some embodiments, tile support is to be configured. For example, how many rows and columns per tile, tiles that are to be used, etc. are configured. A TILECONFIG instruction is an improvement to a computer itself as it provides for support to configure the computer to use a matrix accelerator (either as a part of a processor core, or as an external device). In particular, an execution of the TILECONFIG instruction causes a configuration to be retrieved from memory and applied to matrix (tile) settings within a matrix accelerator.
I. Tile Usage Configuration
Execution circuitry 1811 of a processor/core 1805 stores aspects of tile description 1803 into tile configurations 1817. Tile configurations 1817 detail what tiles for a palette are configured (the number of rows and columns in each tile) and a marking that matrix support is in use. Execution circuitry 1811 is configured to use tiles as specified by tile configurations 1817. Execution circuitry 1811 may also include a machine state register, model-specific register (MSR), or configuration register to indicate tile usage. Additional values such as in-use and start values are also set. Tile configurations 1817 utilize one or more registers 1819 to store tile usage and configuration information.
Byte 1 stores a value to be stored in a “startRow” register 1903 and byte 2 stores a value to be stored in a “startP” register 1905. To support restarting instructions after these events, the instructions store information these registers. To support restarting instructions after break events such as those detailed above, the instructions store information in these registers. The startRow value indicates the row that should be used for restart. The startP value indicates the position within the row for store operations when pairs are used and, in some embodiments, indicates the lower half of the row (in the lower tile of a pair) or higher half of the row (in the higher tile of a pair). Generally, the position in the row (the column) is not needed.
With the exception of TILECONFIG and STTILECFG, successfully executing matrix (tile) instructions will set both startRow and startP to zero.
Any time an interrupted matrix (tile) instruction is not restarted, it is the responsibility of software to zero the startRow and startP values. For example, unmasked floating point exception handlers might decide to finish the operation in software and change the program counter value to another instruction, usually the next instruction. In this case the software exception handler must zero the startRow and startP values in the exception presented to it by the operating system before resuming the program. The operating system will subsequently reload those values using a restore instruction.
Byte 3 stores an indication of pairs (1 b per tile) of tiles 1907.
Bytes 16-17 store the number of rows 1913 and columns 1915 for tile 0, bytes 18-19 store the number of rows and columns for tile 1, etc. In other words, each 2 byte group specifies a number of rows and columns for a tile. If a group of 2 bytes is not used to specify tile parameters, they should have the value zero. Specifying tile parameters for more tiles than the implementation limit or the palette limit results in a fault. Unconfigured tiles are set to an initial state with 0 rows, 0 columns.
Finally, the configuration in memory typically ends with an ending delineation such as all zeros for several consecutive bytes.
II. Exemplary Tile and Tile Configuration Storage
Other combinations are contemplated such as combining the start registers into a single register where they are shown separately, etc.
Chained Tile Operations
Single-instruction, multiple data (SIMD) processor cores implemented according to disclosed embodiments capture significant data parallelism through SIMD execution. However, there is sometimes a limitation to SIMD compute throughput, as the throughput is limited by the number of SIMD functional units or the width of SIMD functional units. SIMD functional units are sometimes referred to herein as processing engines (PEs). Disclosed embodiments leverage the TILES architecture described herein for use in SIMD execution, allowing an increased number of SIMD functional units, and an increased width of each functional unit.
Disclosed embodiments provide a boost in energy efficiency and execution speed for many vectorizable instructions using the TILES architecture to enable reuse of data among chained tile operations accessing tiles within the array. As used herein, a chain is a sequence of two or more instructions (macro operations and/or micro operations) that have source and destination tile specifiers in common. As used herein, a source tile specifier is included in an instruction to specify a location from where to retrieve a source tile to be operated on. A destination tile specifier is included in an instruction to specify a destination tile to which to write a result of the instruction. Specified source and destination tiles may be stored in any one or more of a memory, a cache hierarchy, a scratchpad, and a register file.
To describe a non-limiting example of a chain of tile operations, a first instruction can produce a result that is to be written to a specified destination tile, but the result can also be used as a specified source tile of a second instruction. According to some embodiments, the destination tile specifier of the first instruction is to be set aside, and the result of the first instruction is instead to be routed to a processing engine executing the second instruction. In some other embodiments, the destination tile specifier of the first instruction is to be ignored, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be contravened, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be disregarded, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be disobeyed, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be rejected, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain.
In some embodiments, the result of the first instruction is rerouted to multiple subsequent instructions. In some embodiments, the second instructions is to receive results of multiple prior instructions in the chain.
Chains of instructions are not limited in their length; and many instructions may be chained. A chain of instructions is further described and illustrated at least with respect to
To improve processor efficiency by enabling reuse of data among multiple operations when executing a chain of instructions, disclosed embodiments improve efficiency by opting to set aside a first instruction's destination tile specifier, and to instead route a result of the first instruction directly to a second instruction, without first writing the result back to memory. Such setting aside can be implemented for multiple instructions in the chain, without limitation.
Ignoring the destination tile specifier of the first instruction improves performance, at least because the second instruction need not wait for results of the first instruction to be written back to storage. In some embodiments, instructions within a chain of instructions include a chain control field to provide a hint to execution circuitry as to whether and how to perform the setting aside. In some embodiments, execution circuitry includes a binary translator that monitors dependencies among multiple instructions and determines whether and when to set aside one or more destination tile specifiers and to instead route a results of one of the multiple instructions to one or more PEs executing next instructions in the chain of instructions.
In some embodiments, execution circuitry conducts regular consistency checks to determine any problems caused by the setting aside. In some embodiments, execution circuitry, maintains a record of the instructions whose destination tile specifiers were set aside, so that the SIMD processor can roll back and re-execute those recorded instructions, this time without the setting aside.
Disclosed embodiments do not attempt to increase SIMD efficiency by limiting the choice of execution cores to one type. Rather, some embodiments utilize a mix of asymmetric processing engines (PEs), some having more limited functionality and utilizing less power than others. In some embodiments, a SIMD execution path along with the selection of processing engines from among asymmetric processing engines is done dynamically at runtime.
Some disclosed embodiments take advantage of a paradigm of an array of asymmetric processing engines (PEs). In some embodiments, each of the PEs in the asymmetric array is an arithmetic-logic unit (ALU) that can perform only a limited number of operations. For example, the array of asymmetric PEs may include some PEs dedicated to shifting, some PEs dedicated to addition and subtraction, some PEs dedicated to format conversion, some PEs dedicated to division, some PEs dedicated to multiplication, some PEs dedicated to unary operations, some PEs dedicated to logical operations, and so on, without limitation. In some embodiments, at least some of the PEs are capable of executing more than one type of operation. In operation, execution circuitry is to dynamically select and configure a set of PEs to perform a set of instructions in a chain of SIMD instructions. Different PEs can be selected to perform different operations, matching performance required to performance provided, thereby yielding a power-efficient implementation. One advantage of this asymmetric array is that it allows a close match between processing power required and power expended: only as much processing power as needed is used. In some embodiments, power is conserved by executing a one-time set-up and configuration of a set of PEs to perform different parts of a chain of tile operations and allowing that configuration to be reused for remaining matrix elements. As described herein, a “tile” may be referred to as, and is a special case of a matrix.
In some embodiments, the asymmetric array of PEs includes a dynamically configurable interconnect to route data among multiple PEs in a SIMD processing path. In operation, execution circuitry can select a set of PEs to execute a SIMD instruction chain, and dynamically set up a network to transfer data among them. One advantage of this dynamically configurable interconnect is to allow connections among a wide set of PEs without incurring the cost and die area of necessarily connecting every PE to every other PE.
One advantage of this asymmetric array of PEs and dynamically configurable interconnect is that intermediate results need not be written back to a register file or put onto a bypass network; rather, results of one PE can be routed to the next PE in the SIMD processing chain. This hardware thus can be yield a better performing, more energy-efficient SIMD processor.
Exemplary Execution
To optimize the instruction chain according to embodiments disclosed herein, one or more of the instructions' destination tile specifiers are to be set aside, and the results of that instruction are instead to be routed to a subsequent instruction in the chain. In some embodiments, execution circuitry maintains a record of instructions having had their destination tile specifiers set aside and rolls back to recover by re-executing one or more of the recorded instructions if a problem is detected resulting from the setting aside.
In some embodiments, execution circuitry is to maintain a record of instructions having had their destination tile specifiers set aside, so that execution circuitry can roll back and re-execute the instruction again without the setting aside. In some embodiments, execution circuitry is further to save a machine state before the setting aside, and to subsequently execute a machine state consistency check to detect any inconsistencies caused by the setting aside. In some embodiments, execution circuitry, further to setting aside a destination tile specifier of an instruction, is to set a dirty bit in the configuration of that tile, and to detect an inconsistency if a dirty tile is read by a subsequent instruction. Execution circuitry in some embodiments is to generate a fault if such inconsistencies are detected, and in some embodiments, execution circuitry is to roll back and re-execute one or more instructions.
In operation, as shown, instruction chain 2100 is to be fetched from storage as a sequence of instructions (e.g., micro-ops) or macro-instructions, each having a destination tile specifier, and a source tile specifier. Typically, the instructions in the instruction chain specify, as sources and destinations, intermediate tiles that are stored in the scratchpad 2102. In some embodiments, instructions in the instruction chain specify source tiles and/or destination tiles in a register file, in a cache, or in some other memory structure. In some embodiments, a scratchpad is a memory, including one or more registers or latches or even cache memory and other static memory, that is placed relatively proximate to execution circuitry, and stores intermediate data results. At some time, the tiles in the scratchpad are to be committed to memory.
Assuming matrices (tiles) A, B, and C have already been loaded into scratchpad 2102, matrices (tiles) A, B, and C are retrieved from scratchpad 2102 at tile A 2104, tile B 2106, and tile C 2108, respectively.
At add 2110, Tile A and Tile B are added, and the result at tile E 2112 is to be written back to tile E in the scratchpad 2102 (or in a register file, in a cache, or in some other memory structure). Tile E 2112 is shown with a dashed border to highlight that it is a potential optimization point. In some embodiments, as disclosed herein and as illustrated and described with respect to
At multiply 2114, the result of Add 2110, or tile E, depending on whether the specified destination tile of add 2110 was set aside, is routed to a processing engine to multiply by a value (e.g., 3) broadcasted 2120 from scratchpad 2102, and the result at tile F 2116 is to be written back to tile F in the scratchpad 2102. Tile F 2116 is shown with a dashed border to highlight that it is a potential optimization point. In some embodiments, as disclosed herein and as illustrated and described with respect to
At operation subtract 2118, the result of multiply 2114, or tile F, depending on whether the specified destination tile of multiply 2114 was set aside, is forwarded to a processing engine as a source from which tile C 2108 is to be subtracted. Operation subtract 2118 represents the last instruction in the chain, so optimization is not available, and a result, tile D 2122, of operation subtract 2118 is to be stored at 2124 back in scratchpad 2102 (or in a register file, in a cache, or in some other memory structure).
To optimize the instruction chain according to some embodiments, one or more instructions' destination tile specifiers is to be set aside, and those instructions' results are instead to be routed to subsequent instructions in the chain.
In operation, as shown, instruction chain 2150 is to be fetched from storage as a sequence of instructions (e.g., micro-ops) or macro-instructions, each specifying an arithmetic opcode, and having a destination tile specifier and at least one source tile specifier. Typically, the instructions in the instruction chain specify as source tiles and destination tiles, intermediate tiles stored in the scratchpad 2152. At some point, tiles in the scratchpad are committed to memory.
Assuming matrices (tiles) A, B, and C have already been loaded into scratchpad 2152, matrices (tiles) A, B, and C are retrieved from scratchpad 2152 at 2154, 2156, and 2158, respectively.
At 2160, Tile A and Tile B are added in response to an instruction specifying tile A and tile B as source tiles, and tile E as a destination tile. In operation, a SIMD processor is to perform the addition on multiple vector elements concurrently and in parallel. Execution circuitry optimizes the flow by setting aside the destination tile specifier at 2160 and instead routing the result of the add to the next instruction in the chain, multiply 2164. In operation, a processor is to perform multiply 2164 on multiple vector elements concurrently and in parallel.
Determining Whether and When to Set Aside
According to embodiments disclosed herein, there are multiple ways that execution circuitry can be triggered to optimize the flow by rerouting the result of operation add 2160 to operation multiply 2164.
First, in some embodiments, execution circuitry includes a binary translator that buffers a sequence of instructions and analyzes them dynamically at runtime to detect data dependencies among instructions. The binary translator, for example, can predict that a result of an executed operation will be used only once, by the next instruction in the chain, and will not be used again. Based on such a prediction, execution circuitry can opt to set aside a specified destination tile in the scratchpad, and instead to route the result to the next processing engine executing the next instruction in the chain.
Moreover, in some disclosed embodiments, the add 2160 instruction is formatted according to a chain-friendly instruction format, such as CHAIN_TILE_OP, which, as illustrated and described further below with respect to
Further ways to determine whether a chain exists and to optimize it are disclosed, without limitation. In some embodiments, the chain control field of a CHAIN_TILE_OP includes a header to identify a first instruction and one or more subsequent instructions as being part of the chain or instructions to be optimized according to disclosed embodiments. In some embodiments, a chain control field is to indicate a chain position hint being, for example, one of a start of chain hint, a middle of chain hint, and an end of chain hint to mark each instruction in the chain. In some embodiments, the chain control field of each instruction in the chain specifies an index indicting the instruction's relative position in the chain.
Further, in some disclosed embodiments, a start-of-chain instruction, as further described and illustrated with respect to
According to disclosed embodiments, execution circuitry, in response to determining existence of a chain of instructions to optimize, is to dynamically select and configure a SIMD path of processing engines to perform the chain of instructions, to set aside the specified destination tiles of all but the last instruction in the chain, and to route results of all but the last instruction to the next PE performing a next instruction in the chain. Accordingly, as shown, execution circuitry determines the existence of a chain and reroutes the result of add 2160 from the specified destination tile, tile E, to the PE executing multiply 2164. Similarly, execution circuitry continues processing the chain by rerouting the result of multiply 2164 from specified destination tile F to the PE executing the next instruction in the chain, subtract 2168. In operation, a SIMD processor is to perform the addition 2160, the multiply 2164, and the subtraction 2168 on multiple tile elements concurrently and in parallel. There being no further instructions in the chain after subtract 2168, execution circuitry routes the result of subtract 2168 to specified destination tile D 2172 to be stored at 2174 back to the scratchpad 2152.
Selecting Processing Engines (PEs) and Configuring a SIMD Path
To implement chained tile operations, disclosed embodiments select processing engines to perform the chained mathematical operations. In some embodiments, a SIMD processor includes an asymmetric array of processing engines (PEs). Such an array has a variety of different types of processing engines, some having more processing capabilities but consuming more power or die area than others.
In some embodiments, the SIMD processor has processing engines matched to mathematical operations being performed. For example, some processing engines may be dedicated to unary operations, such as increment, decrement, square root, negate, or trigonometric operations, such as sine, cosine, and tangent, to name a few, non-limiting examples. Some PEs may be dedicated to binary operations, such as mathematical operations (e.g., Add, Subtract, Divide, Multiply, Modulus, etc.), and logical operations (e.g. AND, OR, NAND, NOR, XOR, XNOR, CMP, etc.), to name a few, non-limiting examples. In some embodiments, the SIMD processor has a pool of symmetric processing engines for use in implementing chained tile operations.
In operation, according to some embodiments, execution circuitry selects and configures PEs to execute chained tile operations. For example, in the embodiment of
Some disclosed embodiments take advantage of a paradigm of an array of asymmetric processing engines (PEs). In some embodiments, each of the PEs in the asymmetric array is an arithmetic-logic unit (ALU) that can perform only a limited number of operations. For example, the array of asymmetric PEs may include some PEs dedicated to shifting, some PEs dedicated to addition and subtraction, some PEs dedicated to format conversion, some PEs dedicated to division, some PEs dedicated to multiplication, some PEs dedicated to unary operations, some PEs dedicated to logical operations, and so on, without limitation. In some embodiments, at least some of the PEs are capable of executing more than one type of operation. In operation, execution circuitry is to dynamically select and configure a set of PEs to perform a set of instructions in a chain of SIMD instructions. Different PEs can be selected to perform different operations, matching performance required to performance provided, thereby yielding a power-efficient implementation. One advantage of this asymmetric array is that it allows a close match between processing power required and power expended: only as much processing power as needed is used. In some embodiments, power is conserved by executing a one-time set-up and configuration of a set of PEs to perform different parts of a chain of tile operations and allowing that configuration to be reused for remaining matrix elements. As described herein, a “tile” may be referred to as, and is a special case of a matrix.
Exemplary Processor Execution Flows
At 2211, execution circuitry of the processor is to identify first and second decoded instructions belonging to a chain of instructions. The section entitled “Determining whether and when to set aside,” describes various ways the processor is to identify the first and second decoded instructions belonging to a chain of instruction. One embodiment, for example, entails the processor fetching a start-of-chain instruction before fetching the plurality of instructions at 2201, as further described and illustrated with respect to
At 2213, execution circuitry of the processor is to dynamically select and configure a SIMD path comprising first and second processing engines (PE) to execute the first and second decoded instructions. At 2215, execution circuitry of the processor is to dynamically configure a network path from the first PE to the second PE. Operation 2215 is optional, as indicated by its dashed border, insofar as the network path may already have been configured.
At 2217, according to the illustrated embodiment, execution circuitry of the processor is to set aside the destination tile specifier of the first decoded instruction, and to instead route a result of the first decoded instruction from the first PE to the second PE to be used by the second PE to perform the second decoded instruction. In some other embodiments, the destination tile specifier of the first instruction is to be ignored, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be contravened, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be disregarded, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be disobeyed, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain. In some other embodiments, the destination tile specifier of the first instruction is to be rejected, and the result of the first instruction is instead to be routed to a processing engine executing a subsequent instruction in the chain.
In some embodiments, execution circuitry is to maintain (not shown in
At 2219, the processor is to commit a result of the executed instructions. Operation 2219 is optional, as indicated by its dashed borders, insofar as a writeback of results may occur at a different time, or not at all.
At 2302, a processor is to fetch, decode, and buffer one or more instructions until a set of instructions has been fetched, each instruction specifying source and destination matrices (tiles).
At 2304, the processor tests for a number of error conditions, the presence of any of which is to generate a fault at 2306. As shown, the error conditions tested for at 2304 include: 1) whether tiles are not yet configured, 2) whether the destination and two source tiles are valid, 3) whether either of the two sources is a pair, and 4) any of the various numbers of rows and columns is out of range.
When none of the fault conditions tested at 2304 is true, the processor at 2308 determines whether first and second decoded instructions are part of a chain of instructions. If not, the processor returns to 2302 to fetch at least one more instruction until the set of instructions has been fetched. But if so, the processor at 2310 dynamically selects and configures a SIMD execution path including first and second processing engines (PEs) to execute the first and second decoded instructions. At 2312, the processor is to set aside the destination tile specifier of the first decoded instruction, and instead route a result of the first decoded instruction from the first PE to the second PE to be used by the second PE when executing the second decoded instruction. At 2314, the processor determines whether the chain is complete. If not, the processor returns to 2302 to fetch at least one more instruction until the set of instructions has been fetched. But if so, flow 2300 ends.
IV. Exemplary Pseudocode
Pseudocode in
II. Exemplary Instruction Format(S)
Opcode 2501 is to specify an arithmetic tile operation to be performed on every element of a tile. The specified operation may be a unary operation, a binary operation (in which case source 22505 is also to be specified), or a ternary operation (in which case source 22505 and source 32506 are to be specified). Examples of unary operations to be specified include increment, decrement, square root, negate, or trigonometric operations, such as sine, cosine, and tangent, to name a few, non-limiting examples. Examples of binary operations to be specified include a few examples of which, without limitation, include mathematical operations (e.g., Add, Subtract, Divide, Multiply, Modulus, etc.), logical operations (e.g. AND, OR, NAND, NOR, XOR, XNOR, CMP, etc.), exchange, threshold, and floor, to name a few, non-limiting examples.
In some embodiments, chain control 2502, as illustrated and described at least with respect to
Destination 2503, source 12504, source 22505, and source 32506 are tile specifiers to specify a destination tile and source tiles. Destination 2503, in some embodiments, is a destination tile specifier to identify a destination of a chained tile instruction. In some embodiments, destination 2503 is a destination tile specifier that specifies two or more destinations for a result of the instruction.
Immediate 2507 is to allow for the specification of an immediate in an instruction. Immediate 2507 is optional, as indicated by its dashed border, in the sense that it is not present in an implementation of the generic vector friendly format that does not support an immediate and it is not present in instructions that do not use an immediate.
Writemask 2508 is an optional field (as indicated by its dashed border) whose content controls, on a per data element position basis, whether that data element position in the destination tile is to reflect the result of the operation. Some instructions support merging-writemasking, while others support both merging- and zeroing-writemasking. When merging, vector masks allow any set of elements in the destination tile to be protected from updates during the execution of any operation; in another embodiment, preserving the old value of each element of the destination tile where the corresponding mask bit has a 0. In contrast, zeroing vector masks allow any set of elements in the destination to be zeroed during the execution of any operation. If no writemask is specified, all destination elements are treated as unmasked.
Size 2509 is an optional field (as indicated by its dashed border) to indicate the size of tile data elements being operated on. Size 2509 is shown as a separate instruction operand, but the asterisk is included to indicate that in some embodiments, it is to be included in an opcode, prefix or suffix, “B,” “W,” “D,” and “Q,” corresponding to a size-1 byte, 2 bytes, 4 bytes, or 8 bytes, respectively of each tile element. In some embodiments, size 2509 is included in the opcode, such as a prefix or suffix, “H,” “S,” “D,” “Q,” and “E,” corresponding to precision levels—Half precision (2 bytes), Single precision (4 bytes), Double precision (8 bytes), and Quadruple precision (16 byes), respectively—of each tile element.
Opcode 2511 causes a processor executing the instruction to recognize the instruction as a start-of-chain instruction and may include prefixes or suffixes to further specify the behavior of the instruction. For example, opcode 2511 may include a prefix to specify that the instruction includes fields to specify chain targets 2513, secondary chain targets 2514, and chain ops 2515. In operation, a processor, according to some embodiments, in response to the start-of-chain instruction 2510 is to select and configure a SIMD path of processing engines to execute the chain of instructions.
Chain length 2512 in some embodiments is an immediate or a constant to specify a number of instructions in the chain of instructions. In some embodiments, chain length 2512 specifies a general purpose register whose contents specify the number of instructions in the chain of instructions.
Chain targets 2513 is to specify a target of at least the first instruction in the chain of instructions. As shown, chain targets 2513 is to specify targets 2513A-2513H for each of the first seven instructions in a chain of instructions having a length of eight (The result of the eight instruction in the chain is to be written to the specified destination, rather than to be rerouted to a next instruction). In some embodiments, the chain targets 2513 field is an immediate, logically partitioned into bits to specify the targets of each of the instructions in the chain. In some embodiments, chain targets 2513 specifies a general purpose register containing the target information. In some embodiments, chain targets 2513 specifies a vector register containing the target information. In some embodiments, chain targets 2513 specifies a model-specific register (“MSR”) where the target information is stored. In other embodiments, chain targets 2513 specifies a memory location where the target information is stored.
In some embodiments, a 4-byte chain targets 2513 field may be partitioned into 4-bit nibbles to specify targets for each of up to 8 instructions in a chain of instructions. In some embodiments, each chain target 2513A-2513H specifies an absolute position of a target instruction in the chain. As shown, target 12513A (of instruction 0) is to specify instruction 1; target 225138 is to specify instruction 4; target 32513C; target 42513D is to specify instruction 4; target 52513E is to specify instruction 6; target 62513F is to specify instruction 6; and, target 72513G is to specify instruction 7.
In some embodiments, each of targets 1-7, 2513A-G, specifies a relative offset of a target instruction to which to route a result. Each of targets 2513A-G could be a 2-bit or 3-bit value specifying a distance between an instruction and its target. For example, as shown, target 12513A to target 72513G could specify 1, 3, 2, 1, 2, 1, 1, respectively.
In some embodiments, secondary chain targets 2514 are formatted like chain targets 2513 but specify a second target for each of the instructions in the chain of instructions. As with chain targets 2513, secondary chain targets 2514 can be an immediate, a general purpose register, a vector register, a model-specific register, or a memory location. Secondary chain targets 2514 is optional, as indicated by its dashed border.
In some embodiments, start-of-chain instruction 2510 includes chain ops 2515 to specify the operations to be performed by each of the chain of instructions. In operation, a processor is to use the chain ops 2515 field to inform its selection, configuration, and routing of processing engines along the SIMD execution path.
In some embodiments, encodings of the instruction include a scale-index-base (SIB) type memory addressing operand that indirectly identifies multiple indexed destination locations in memory. In one embodiment, an SIB type memory operand may include an encoding identifying a base address register. The contents of the base address register may represent a base address in memory from which the addresses of the particular destination locations in memory are calculated. For example, the base address may be the address of the first location in a block of potential destination locations for an extended vector instruction. In one embodiment, a SIB type memory operand may include an encoding identifying an index register. Each element of the index register may specify an index or offset value usable to compute, from the base address, an address of a respective destination location within a block of potential destination locations. In one embodiment, a SIB type memory operand may include an encoding specifying a scaling factor to be applied to each index value when computing a respective destination address. For example, if a scaling factor value of four is encoded in the SIB type memory operand, each index value obtained from an element of the index register may be multiplied by four and then added to the base address to compute a destination address.
In one embodiment, an SIB type memory operand of the form vm32{x, y, z} may identify a vector array of memory operands specified using SIB type memory addressing. In this example, the array of memory addresses is specified using a common base register, a constant scaling factor, and a vector index register containing individual elements, each of which is a 32-bit index value. The vector index register may be a 128-bit (e.g., XMM) register (vm32x), a 256-bit (e.g., YMM) register (vm32y), or a 512-bit (e.g., ZMM) register (vm32z). In another embodiment, an SIB type memory operand of the form vm64{x, y, z} may identify a vector array of memory operands specified using SIB type memory addressing. In this example, the array of memory addresses is specified using a common base register, a constant scaling factor, and a vector index register containing individual elements, each of which is a 64-bit index value. The vector index register may be a 128-bit register (e.g., XMM) register (vm64x), a 256-bit (e.g., YMM) register (vm64y) or a 512-bit (e.g., ZMM) register (vm64z).
V. Detailed Exemplary Systems, Processors, and Emulation
Detailed herein are examples of hardware, software, etc. to execute the above described instructions. For example, what is described below details aspects of instruction execution including various pipeline stages such as fetch, decode, schedule, execute, retire, etc.
Instruction Sets
An instruction set may include one or more instruction formats. A given instruction format may define various fields (e.g., number of bits, location of bits) to specify, among other things, the operation to be performed (e.g., opcode) and the operand(s) on which that operation is to be performed and/or other data field(s) (e.g., mask). Some instruction formats are further broken down though the definition of instruction templates (or subformats). For example, the instruction templates of a given instruction format may be defined to have different subsets of the instruction format's fields (the included fields are typically in the same order, but at least some have different bit positions because there are less fields included) and/or defined to have a given field interpreted differently. Thus, each instruction of an ISA is expressed using a given instruction format (and, if defined, in a given one of the instruction templates of that instruction format) and includes fields for specifying the operation and the operands. For example, an exemplary ADD instruction has a specific opcode and an instruction format that includes an opcode field to specify that opcode and operand fields to select operands (source1/destination and source2); and an occurrence of this ADD instruction in an instruction stream will have specific contents in the operand fields that select specific operands. A set of SIMD extensions referred to as the Advanced Vector Extensions (AVX) (AVX1 and AVX2) and using the Vector Extensions (VEX) coding scheme has been released and/or published (e.g., see Intel® 64 and IA-32 Architectures Software Developer's Manual, September 2014; and see Intel® Advanced Vector Extensions Programming Reference, October 2014).
Exemplary Instruction Formats
Embodiments of the instruction(s) described herein may be embodied in different formats. Additionally, exemplary systems, architectures, and pipelines are detailed below. Embodiments of the instruction(s) may be executed on such systems, architectures, and pipelines, but are not limited to those detailed.
Generic Vector Friendly Instruction Format
A vector friendly instruction format is an instruction format that is suited for vector instructions (e.g., there are certain fields specific to vector operations). While embodiments are described in which both vector and scalar operations are supported through the vector friendly instruction format, alternative embodiments use only vector operations the vector friendly instruction format.
While embodiments of the invention will be described in which the vector friendly instruction format supports the following: a 64 byte vector operand length (or size) with 32 bit (4 byte) or 64 bit (8 byte) data element widths (or sizes) (and thus, a 64 byte vector consists of either 16 doubleword-size elements or alternatively, 8 quadword-size elements); a 64 byte vector operand length (or size) with 16 bit (2 byte) or 8 bit (1 byte) data element widths (or sizes); a 32 byte vector operand length (or size) with 32 bit (4 byte), 64 bit (8 byte), 16 bit (2 byte), or 8 bit (1 byte) data element widths (or sizes); and a 16 byte vector operand length (or size) with 32 bit (4 byte), 64 bit (8 byte), 16 bit (2 byte), or 8 bit (1 byte) data element widths (or sizes); alternative embodiments may support more, less and/or different vector operand sizes (e.g., 256 byte vector operands) with more, less, or different data element widths (e.g., 128 bit (16 byte) data element widths).
The class A instruction templates in
The generic vector friendly instruction format 2516 includes the following fields listed below in the order illustrated in
Format field 2540—a specific value (an instruction format identifier value) in this field uniquely identifies the vector friendly instruction format, and thus occurrences of instructions in the vector friendly instruction format in instruction streams. As such, this field is optional in the sense that it is not needed for an instruction set that has only the generic vector friendly instruction format.
Base operation field 2542—its content distinguishes different base operations.
Register index field 2544—its content, directly or through address generation, specifies the locations of the source and destination operands, be they in registers or in memory. These include a sufficient number of bits to select N registers from a P×Q (e.g. 32×512, 16×128, 32×1024, 64×1024) register file. While in one embodiment N may be up to three sources and one destination register, alternative embodiments may support more or less sources and destination registers (e.g., may support up to two sources where one of these sources also acts as the destination, may support up to three sources where one of these sources also acts as the destination, may support up to two sources and one destination).
Modifier field 2546—its content distinguishes occurrences of instructions in the generic vector instruction format that specify memory access from those that do not; that is, between no memory access 2519 instruction templates and memory access 2520 instruction templates. Memory access operations read and/or write to the memory hierarchy (in some cases specifying the source and/or destination addresses using values in registers), while non-memory access operations do not (e.g., the source and destinations are registers). While in one embodiment this field also selects between three different ways to perform memory address calculations, alternative embodiments may support more, less, or different ways to perform memory address calculations.
Augmentation operation field 2550—its content distinguishes which one of a variety of different operations to be performed in addition to the base operation. This field is context specific. In one embodiment of the invention, this field is divided into a class field 2568, an alpha field 2552, and a beta field 2554. The augmentation operation field 2550 allows common groups of operations to be performed in a single instruction rather than 2, 3, or 4 instructions.
Scale field 2560—its content allows for the scaling of the index field's content for memory address generation (e.g., for address generation that uses 2scale*index+base).
Displacement Field 2562A—its content is used as part of memory address generation (e.g., for address generation that uses 2scale*index+base+displacement).
Displacement Factor Field 25628 (note that the juxtaposition of displacement field 2562A directly over displacement factor field 25628 indicates one or the other is used)—its content is used as part of address generation; it specifies a displacement factor that is to be scaled by the size of a memory access (N)—where N is the number of bytes in the memory access (e.g., for address generation that uses 2scale*index+base+scaled displacement). Redundant low-order bits are ignored and hence, the displacement factor field's content is multiplied by the memory operands total size (N) in order to generate the final displacement to be used in calculating an effective address. The value of N is determined by the processor hardware at runtime based on the full opcode field 2574 (described later herein) and the data manipulation field 2554C. The displacement field 2562A and the displacement factor field 2562B are optional in the sense that they are not used for the no memory access 2519 instruction templates and/or different embodiments may implement only one or none of the two.
Data element width field 2564—its content distinguishes which one of a number of data element widths is to be used (in some embodiments for all instructions; in other embodiments for only some of the instructions). This field is optional in the sense that it is not needed if only one data element width is supported and/or data element widths are supported using some aspect of the opcodes.
Write mask field 2570—its content controls, on a per data element position basis, whether that data element position in the destination vector operand reflects the result of the base operation and augmentation operation. Class A instruction templates support merging-writemasking, while class B instruction templates support both merging- and zeroing-writemasking. When merging, vector masks allow any set of elements in the destination to be protected from updates during the execution of any operation (specified by the base operation and the augmentation operation); in other one embodiment, preserving the old value of each element of the destination where the corresponding mask bit has a 0. In contrast, when zeroing vector masks allow any set of elements in the destination to be zeroed during the execution of any operation (specified by the base operation and the augmentation operation); in one embodiment, an element of the destination is set to 0 when the corresponding mask bit has a 0 value. A subset of this functionality is the ability to control the vector length of the operation being performed (that is, the span of elements being modified, from the first to the last one); however, it is not necessary that the elements that are modified be consecutive. Thus, the write mask field 2570 allows for partial vector operations, including loads, stores, arithmetic, logical, etc. While embodiments of the invention are described in which the write mask field's 2570 content selects one of a number of write mask registers that contains the write mask to be used (and thus the write mask field's 2570 content indirectly identifies that masking to be performed), alternative embodiments instead or additional allow the write mask field's 2570 content to directly specify the masking to be performed.
Immediate field 2572—its content allows for the specification of an immediate. This field is optional in the sense that it is not present in an implementation of the generic vector friendly format that does not support immediate and it is not present in instructions that do not use an immediate.
Class field 2568—its content distinguishes between different classes of instructions. With reference to
Instruction Templates of Class A
In the case of the no memory access 2519 instruction templates of class A, the alpha field 2552 is interpreted as an RS field 2552A, whose content distinguishes which one of the different augmentation operation types are to be performed (e.g., round 2552A.1 and data transform 2552A.2 are respectively specified for the no memory access, full round type operation 2517 and the no memory access, data transform type operation 2518 instruction templates), while the beta field 2554 distinguishes which of the operations of the specified type is to be performed. In the no memory access 2519 instruction templates, the scale field 2560, the displacement field 2562A, and the displacement scale filed 2562B are not present.
No-Memory Access Instruction Templates—Full Round Control Type Operation
In the no memory access full round control type operation 2517 instruction template, the beta field 2554 is interpreted as a round control field 2554A, whose content(s) provide static rounding. While in the described embodiments of the invention the round control field 2554A includes a suppress all floating point exceptions (SAE) field 2556 and a round operation control field 2558, alternative embodiments may encode both these concepts into the same field or only have one or the other of these concepts/fields (e.g., may have only the round operation control field 2558).
SAE field 2556—its content distinguishes whether or not to disable the exception event reporting; when the SAE field's 2556 content indicates suppression is enabled, a given instruction does not report any kind of floating-point exception flag and does not raise any floating point exception handler.
Round operation control field 2558—its content distinguishes which one of a group of rounding operations to perform (e.g., Round-up, Round-down, Round-towards-zero and Round-to-nearest). Thus, the round operation control field 2558 allows for the changing of the rounding mode on a per instruction basis. In one embodiment of the invention where a processor includes a control register for specifying rounding modes, the content of round operation control field 2558 overrides that register value.
No Memory Access Instruction Templates—Data Transform Type Operation
In the no memory access data transform type operation 2518 instruction template, the beta field 2554 is interpreted as a data transform field 25546, whose content distinguishes which one of a number of data transforms is to be performed (e.g., no data transform, swizzle, broadcast).
In the case of a memory access 2520 instruction template of class A, the alpha field 2552 is interpreted as an eviction hint field 2552B, whose content distinguishes which one of the eviction hints is to be used (in
Vector memory instructions perform vector loads from and vector stores to memory, with conversion support. As with regular vector instructions, vector memory instructions transfer data from/to memory in a data element-wise fashion, with the elements that are actually transferred is dictated by the contents of the vector mask that is selected as the write mask.
Memory Access Instruction Templates—Temporal
Temporal data is data likely to be reused soon enough to benefit from caching. This is, however, a hint, and different processors may implement it in different ways, including ignoring the hint entirely.
Memory Access Instruction Templates—Non-Temporal
Non-temporal data is data unlikely to be reused soon enough to benefit from caching in the 1st-level cache and should be given priority for eviction. This is, however, a hint, and different processors may implement it in different ways, including ignoring the hint entirely.
Instruction Templates of Class B
In the case of the instruction templates of class B, the alpha field 2552 is interpreted as a write mask control (Z) field 2552C, whose content distinguishes whether the write masking controlled by the write mask field 2570 should be a merging or a zeroing.
In the case of the no memory access 2519 instruction templates of class B, part of the beta field 2554 is interpreted as an RL field 2557A, whose content distinguishes which one of the different augmentation operation types are to be performed (e.g., round 2557A.1 and vector length (VSIZE) 2557A.2 are respectively specified for the no memory access, write mask control, partial round control type operation 2522 instruction template and the no memory access, write mask control, VSIZE type operation 2523 instruction template), while the rest of the beta field 2554 distinguishes which of the operations of the specified type is to be performed. In the no memory access 2519 instruction templates, the scale field 2560, the displacement field 2562A, and the displacement scale filed 2562B are not present.
In the no memory access, write mask control, partial round control type operation 2522 instruction template, the rest of the beta field 2554 is interpreted as a round operation control field 2559A and exception event reporting is disabled (a given instruction does not report any kind of floating-point exception flag and does not raise any floating point exception handler).
Round operation control field 2559A—just as round operation control field 2558, its content distinguishes which one of a group of rounding operations to perform (e.g., Round-up, Round-down, Round-towards-zero and Round-to-nearest). Thus, the round operation control field 2559A allows for the changing of the rounding mode on a per instruction basis. In one embodiment of the invention where a processor includes a control register for specifying rounding modes, the content of round operation control field 2559A overrides that register value.
In the no memory access, write mask control, VSIZE type operation 2523 instruction template, the rest of the beta field 2554 is interpreted as a vector length field 2559B, whose content distinguishes which one of a number of data vector lengths is to be performed on (e.g., 128, 256, or 512 byte).
In the case of a memory access 2520 instruction template of class B, part of the beta field 2554 is interpreted as a broadcast field 2557B, whose content distinguishes whether or not the broadcast type data manipulation operation is to be performed, while the rest of the beta field 2554 is interpreted the vector length field 2559B. The memory access 2520 instruction templates include the scale field 2560, and optionally the displacement field 2562A or the displacement scale field 2562B.
With regard to the generic vector friendly instruction format 2516, a full opcode field 2574 is shown including the format field 2540, the base operation field 2542, and the data element width field 2564. While one embodiment is shown where the full opcode field 2574 includes all of these fields, the full opcode field 2574 includes less than all of these fields in embodiments that do not support all of them. The full opcode field 2574 provides the operation code (opcode).
The augmentation operation field 2550, the data element width field 2564, and the write mask field 2570 allow these features to be specified on a per instruction basis in the generic vector friendly instruction format.
The combination of write mask field and data element width field create typed instructions in that they allow the mask to be applied based on different data element widths.
The various instruction templates found within class A and class B are beneficial in different situations. In some embodiments of the invention, different processors or different cores within a processor may support only class A, only class B, or both classes. For instance, a high performance general purpose out-of-order core intended for general-purpose computing may support only class B, a core intended primarily for graphics and/or scientific (throughput) computing may support only class A, and a core intended for both may support both (of course, a core that has some mix of templates and instructions from both classes but not all templates and instructions from both classes is within the purview of the invention). Also, a single processor may include multiple cores, all of which support the same class or in which different cores support different class. For instance, in a processor with separate graphics and general purpose cores, one of the graphics cores intended primarily for graphics and/or scientific computing may support only class A, while one or more of the general purpose cores may be high performance general purpose cores with out of order execution and register renaming intended for general-purpose computing that support only class B. Another processor that does not have a separate graphics core, may include one more general purpose in-order or out-of-order cores that support both class A and class B. Of course, features from one class may also be implement in the other class in different embodiments of the invention. Programs written in a high level language would be put (e.g., just in time compiled or statically compiled) into an variety of different executable forms, including: 1) a form having only instructions of the class(es) supported by the target processor for execution; or 2) a form having alternative routines written using different combinations of the instructions of all classes and having control flow code that selects the routines to execute based on the instructions supported by the processor which is currently executing the code.
Exemplary Specific Vector Friendly Instruction Format
It should be understood that, although embodiments of the invention are described with reference to the specific vector friendly instruction format 2600 in the context of the generic vector friendly instruction format 2516 for illustrative purposes, the invention is not limited to the specific vector friendly instruction format 2600 except where claimed. For example, the generic vector friendly instruction format 2516 contemplates a variety of possible sizes for the various fields, while the specific vector friendly instruction format 2600 is shown as having fields of specific sizes. By way of specific example, while the data element width field 2564 is illustrated as a one bit field in the specific vector friendly instruction format 2600, the invention is not so limited (that is, the generic vector friendly instruction format 2516 contemplates other sizes of the data element width field 2564).
The specific vector friendly instruction format 2600 includes the following fields listed below in the order illustrated in
EVEX Prefix (Bytes 0-3) 2602—is encoded in a four-byte form.
Format field 2540 (EVEX Byte 0, bits [7:0])—the first byte (EVEX Byte 0) is the format field 2540 and it contains 0x62 (the unique value used for distinguishing the vector friendly instruction format in one embodiment of the invention).
The second-fourth bytes (EVEX Bytes 1-3) include a number of bit fields providing specific capability.
REX field 2605 (EVEX Byte 1, bits [7-5])—consists of a EVEX.R bit field (EVEX Byte 1, bit [7]-R), EVEX.X bit field (EVEX byte 1, bit [6]-X), and EVEX.B bit field (EVEX byte 1, bit[5]-B). The EVEX.R, EVEX.X, and EVEX.B bit fields provide the same functionality as the corresponding VEX bit fields, and are encoded using 1s complement form, i.e. ZMM0 is encoded as 1111B, ZMM15 is encoded as 0000B. Other fields of the instructions encode the lower three bits of the register indexes as is known in the art (rrr, xxx, and bbb), so that Rrrr, Xxxx, and Bbbb may be formed by adding EVEX.R, EVEX.X, and EVEX.B.
REX′ 2610A—this is the first part of the REX′ field 2610 and is the EVEX.R′ bit field (EVEX Byte 1, bit [4]-R′) that is used to encode either the upper 16 or lower 16 of the extended 32 register set. In one embodiment of the invention, this bit, along with others as indicated below, is stored in bit inverted format to distinguish (in the well-known x86 32-bit mode) from the BOUND instruction, whose real opcode byte is 62, but does not accept in the MOD R/M field (described below) the value of 11 in the MOD field; alternative embodiments of the invention do not store this and the other indicated bits below in the inverted format. A value of 1 is used to encode the lower 16 registers. In other words, R′Rrrr is formed by combining EVEX.R′, EVEX.R, and the other RRR from other fields.
Opcode map field 2615 (EVEX byte 1, bits [3:0]-mmmm)—its content encodes an implied leading opcode byte (OF, OF 38, or OF 3).
Data element width field 2564 (EVEX byte 2, bit [7]-W)—is represented by the notation EVEX.W. EVEX.W is used to define the granularity (size) of the datatype (either 32-bit data elements or 64-bit data elements).
EVEX.vvvv 2620 (EVEX Byte 2, bits [6:3]-vvvv)—the role of EVEX.vvvv may include the following: 1) EVEX.vvvv encodes the first source register operand, specified in inverted (1s complement) form and is valid for instructions with 2 or more source operands; 2) EVEX.vvvv encodes the destination register operand, specified in 1s complement form for certain vector shifts; or 3) EVEX.vvvv does not encode any operand, the field is reserved and should contain 1111b. Thus, EVEX.vvvv field 2620 encodes the 4 low-order bits of the first source register specifier stored in inverted (1s complement) form. Depending on the instruction, an extra different EVEX bit field is used to extend the specifier size to 32 registers.
EVEX.U 2568 Class field (EVEX byte 2, bit [2]-U)—If EVEX.U=0, it indicates class A or EVEX.U0; if EVEX.0=1, it indicates class B or EVEX.U1.
Prefix encoding field 2625 (EVEX byte 2, bits [1:0]-pp)—provides additional bits for the base operation field. In addition to providing support for the legacy SSE instructions in the EVEX prefix format, this also has the benefit of compacting the SIMD prefix (rather than requiring a byte to express the SIMD prefix, the EVEX prefix requires only 2 bits). In one embodiment, to support legacy SSE instructions that use a SIMD prefix (66H, F2H, F3H) in both the legacy format and in the EVEX prefix format, these legacy SIMD prefixes are encoded into the SIMD prefix encoding field; and at runtime are expanded into the legacy SIMD prefix prior to being provided to the decoder's PLA (so the PLA can execute both the legacy and EVEX format of these legacy instructions without modification). Although newer instructions could use the EVEX prefix encoding field's content directly as an opcode extension, certain embodiments expand in a similar fashion for consistency but allow for different meanings to be specified by these legacy SIMD prefixes. An alternative embodiment may redesign the PLA to support the 2 bit SIMD prefix encodings, and thus not require the expansion.
Alpha field 2552 (EVEX byte 3, bit [7]-EH; also known as EVEX.EH, EVEX.rs, EVEX.RL, EVEX.write mask control, and EVEX.N; also illustrated with α)—as previously described, this field is context specific.
Beta field 2554 (EVEX byte 3, bits [6:4]-SSS, also known as EVEX.s2-0, EVEX.r2-0, EVEX.rr1, EVEX.LL0, EVEX.LLB; also illustrated with βββ)—as previously described, this field is context specific.
REX′ 2610B—this is the remainder of the REX′ field 2610 and is the EVEX.V′ bit field (EVEX Byte 3, bit [3]-V′) that may be used to encode either the upper 16 or lower 16 of the extended 32 register set. This bit is stored in bit inverted format. A value of 1 is used to encode the lower 16 registers. In other words, V′VVVV is formed by combining EVEX.V′, EVEX.vvvv.
Write mask field 2570 (EVEX byte 3, bits [2:0]-kkk)—its content specifies the index of a register in the write mask registers as previously described. In one embodiment of the invention, the specific value EVEX.kkk=000 has a special behavior implying no write mask is used for the particular instruction (this may be implemented in a variety of ways including the use of a write mask hardwired to all ones or hardware that bypasses the masking hardware).
Real Opcode Field 2630 (Byte 4) is also known as the opcode byte. Part of the opcode is specified in this field.
MOD R/M Field 2640 (Byte 5) includes MOD field 2642, Reg field 2644, and R/M field 2646. As previously described, the MOD field's 2642 content distinguishes between memory access and non-memory access operations. The role of Reg field 2644 can be summarized to two situations: encoding either the destination register operand or a source register operand or be treated as an opcode extension and not used to encode any instruction operand. The role of R/M field 2646 may include the following: encoding the instruction operand that references a memory address or encoding either the destination register operand or a source register operand.
Scale, Index, Base (SIB) Byte (Byte 6)—As previously described, the content of scale field 2560 is used for memory address generation. SIB.xxx 2654 and SIB.bbb 2656—the contents of these fields have been previously referred to with regard to the register indexes Xxxx and Bbbb.
Displacement field 2562A (Bytes 7-10)—when MOD field 2642 contains 10, bytes 7-10 are the displacement field 2562A, and it works the same as the legacy 32-bit displacement (disp32) and works at byte granularity.
Displacement factor field 2562B (Byte 7)—when MOD field 2642 contains 01, byte 7 is the displacement factor field 2562B. The location of this field is that same as that of the legacy x86 instruction set 8-bit displacement (disp8), which works at byte granularity. Since disp8 is sign extended, it can only address between −128 and 127 bytes offsets; in terms of 64 byte cache lines, disp8 uses 8 bits that can be set to only four really useful values −128, −64, 0, and 64; since a greater range is often needed, disp32 is used; however, disp32 requires 4 bytes. In contrast to disp8 and disp32, the displacement factor field 2562B is a reinterpretation of disp8; when using displacement factor field 2562B, the actual displacement is determined by the content of the displacement factor field multiplied by the size of the memory operand access (N). This type of displacement is referred to as disp8*N. This reduces the average instruction length (a single byte of used for the displacement but with a much greater range). Such compressed displacement is based on the assumption that the effective displacement is multiple of the granularity of the memory access, and hence, the redundant low-order bits of the address offset do not need to be encoded. In other words, the displacement factor field 2562B substitutes the legacy x86 instruction set 8-bit displacement. Thus, the displacement factor field 2562B is encoded the same way as an x86 instruction set 8-bit displacement (so no changes in the ModRM/SIB encoding rules) with the only exception that disp8 is overloaded to disp8*N. In other words, there are no changes in the encoding rules or encoding lengths but only in the interpretation of the displacement value by hardware (which needs to scale the displacement by the size of the memory operand to obtain a byte-wise address offset). Immediate field 2572 operates as previously described.
Full Opcode Field
Register Index Field
Augmentation Operation Field
When U=1, the alpha field 2552 (EVEX byte 3, bit [7]-EH) is interpreted as the write mask control (Z) field 2552C. When U=1 and the MOD field 2642 contains 11 (signifying a no memory access operation), part of the beta field 2554 (EVEX byte 3, bit [4]-S0) is interpreted as the RL field 2557A; when it contains a 1 (round 2557A.1) the rest of the beta field 2554 (EVEX byte 3, bit [6-5]-S2-1) is interpreted as the round operation control field 2559A, while when the RL field 2557A contains a 0 (VSIZE 2557.A2) the rest of the beta field 2554 (EVEX byte 3, bit [6-5]-S2-1) is interpreted as the vector length field 2559B (EVEX byte 3, bit [6-5]-L1-0). When U=1 and the MOD field 2642 contains 00, 01, or 10 (signifying a memory access operation), the beta field 2554 (EVEX byte 3, bits [6:4]-SSS) is interpreted as the vector length field 2559B (EVEX byte 3, bit [6-5]-L1-0) and the broadcast field 2557B (EVEX byte 3, bit [4]-B).
Exemplary Register Architecture
In other words, the vector length field 2559B selects between a maximum length and one or more other shorter lengths, where each such shorter length is half the length of the preceding length; and instructions templates without the vector length field 2559B operate on the maximum vector length. Further, in one embodiment, the class B instruction templates of the specific vector friendly instruction format 2600 operate on packed or scalar single/double-precision floating point data and packed or scalar integer data. Scalar operations are operations performed on the lowest order data element position in an zmm/ymm/xmm register; the higher order data element positions are either left the same as they were prior to the instruction or zeroed depending on the embodiment.
Write mask registers 2715—in the embodiment illustrated, there are 8 write mask registers (k0 through k7), each 64 bits in size. In an alternate embodiment, the write mask registers 2715 are 16 bits in size. As previously described, in one embodiment of the invention, the vector mask register k0 cannot be used as a write mask; when the encoding that would normally indicate k0 is used for a write mask, it selects a hardwired write mask of 0xFFFF, effectively disabling write masking for that instruction.
General-purpose registers 2725—in the embodiment illustrated, there are sixteen 64-bit general-purpose registers that are used along with the existing x86 addressing modes to address memory operands. These registers are referenced by the names RAX, RBX, RCX, RDX, RBP, RSI, RDI, RSP, and R8 through R15.
Scalar floating point stack register file (x87 stack) 2745, on which is aliased the MMX packed integer flat register file 2750—in the embodiment illustrated, the x87 stack is an eight-element stack used to perform scalar floating-point operations on 32/64/80-bit floating point data using the x87 instruction set extension; while the MMX registers are used to perform operations on 64-bit packed integer data, as well as to hold operands for some operations performed between the MMX and XMM registers.
Alternative embodiments of the invention may use wider or narrower registers. Additionally, alternative embodiments of the invention may use more, less, or different register files and registers.
Exemplary Core Architectures, Processors, and Computer Architectures
Processor cores may be implemented in different ways, for different purposes, and in different processors. For instance, implementations of such cores may include: 1) a general purpose in-order core intended for general-purpose computing; 2) a high performance general purpose out-of-order core intended for general-purpose computing; 3) a special purpose core intended primarily for graphics and/or scientific (throughput) computing. Implementations of different processors may include: 1) a CPU including one or more general purpose in-order cores intended for general-purpose computing and/or one or more general purpose out-of-order cores intended for general-purpose computing; and 2) a coprocessor including one or more special purpose cores intended primarily for graphics and/or scientific (throughput). Such different processors lead to different computer system architectures, which may include: 1) the coprocessor on a separate chip from the CPU; 2) the coprocessor on a separate die in the same package as a CPU; 3) the coprocessor on the same die as a CPU (in which case, such a coprocessor is sometimes referred to as special purpose logic, such as integrated graphics and/or scientific (throughput) logic, or as special purpose cores); and 4) a system on a chip that may include on the same die the described CPU (sometimes referred to as the application core(s) or application processor(s)), the above described coprocessor, and additional functionality. Exemplary core architectures are described next, followed by descriptions of exemplary processors and computer architectures.
Exemplary Core Architectures
In-Order and Out-of-Order Core Block Diagram
In
The front end unit 2830 includes a branch prediction unit 2832 coupled to an instruction cache unit 2834, which is coupled to an instruction translation lookaside buffer (TLB) 2836, which is coupled to an instruction fetch unit 2838, which is coupled to send fetched instruction 2839 to a decode unit 2840. The decode unit 2840 (or decoder) may decode instructions, and generate as an output one or more decoded instructions 2841, or micro-operations, micro-code entry points, microinstructions, other instructions, or other control signals, which are decoded from, or which otherwise reflect, or are derived from, the original instructions. The decode unit 2840 may be implemented using various different mechanisms. Examples of suitable mechanisms include, but are not limited to, look-up tables, hardware implementations, programmable logic arrays (PLAs), microcode read only memories (ROMs), etc. In one embodiment, the core 2890 includes a microcode ROM or other medium that stores microcode for certain macroinstructions (e.g., in decode unit 2840 or otherwise within the front end unit 2830). The decode unit 2840 is coupled to a rename/allocator unit 2852 in the execution engine unit 2850.
The execution engine unit 2850 includes the rename/allocator unit 2852 coupled to a retirement unit 2854 and a set of one or more scheduler unit(s) 2856. The scheduler unit(s) 2856 represents any number of different schedulers, including reservations stations, central instruction window, etc. The scheduler unit(s) 2856 is coupled to the physical register file(s) unit(s) 2858. Each of the physical register file(s) units 2858 represents one or more physical register files, different ones of which store one or more different data types, such as scalar integer, scalar floating point, packed integer, packed floating point, vector integer, vector floating point, status (e.g., an instruction pointer that is the address of the next instruction to be executed), etc. In one embodiment, the physical register file(s) unit 2858 comprises a vector registers unit, a write mask registers unit, and a scalar registers unit. These register units may provide architectural vector registers, vector mask registers, and general purpose registers. The physical register file(s) unit(s) 2858 is overlapped by the retirement unit 2854 to illustrate various ways in which register renaming and out-of-order execution may be implemented (e.g., using a reorder buffer(s) and a retirement register file(s); using a future file(s), a history buffer(s), and a retirement register file(s); using a register maps and a pool of registers; etc.). The retirement unit 2854 and the physical register file(s) unit(s) 2858 are coupled to the execution cluster(s) 2860. The execution cluster(s) 2860 includes a set of one or more execution units 2862 and a set of one or more memory access units 2864. The execution units 2862 may perform various operations (e.g., shifts, addition, subtraction, multiplication) and on various types of data (e.g., scalar floating point, packed integer, packed floating point, vector integer, vector floating point). While some embodiments may include a number of execution units dedicated to specific functions or sets of functions, other embodiments may include only one execution unit or multiple execution units that all perform all functions. The scheduler unit(s) 2856, physical register file(s) unit(s) 2858, and execution cluster(s) 2860 are shown as being possibly plural because certain embodiments create separate pipelines for certain types of data/operations (e.g., a scalar integer pipeline, a scalar floating point/packed integer/packed floating point/vector integer/vector floating point pipeline, and/or a memory access pipeline that each have their own scheduler unit, physical register file(s) unit, and/or execution cluster—and in the case of a separate memory access pipeline, certain embodiments are implemented in which only the execution cluster of this pipeline has the memory access unit(s) 2864). It should also be understood that where separate pipelines are used, one or more of these pipelines may be out-of-order issue/execution and the rest in-order.
The set of memory access units 2864 is coupled to the memory unit 2870, which includes a data TLB unit 2872 coupled to a data cache unit 2874 coupled to a level 2 (L2) cache unit 2876. In one exemplary embodiment, the memory access units 2864 may include a load unit, a store address unit, and a store data unit, each of which is coupled to the data TLB unit 2872 in the memory unit 2870. The instruction cache unit 2834 is further coupled to a level 2 (L2) cache unit 2876 in the memory unit 2870. The L2 cache unit 2876 is coupled to one or more other levels of cache and eventually to a main memory.
By way of example, the exemplary register renaming, out-of-order issue/execution core architecture may implement the pipeline 2800 as follows: 1) the instruction fetch 2838 performs the fetch and length decoding stages 2802 and 2804; 2) the decode unit 2840 performs the decode stage 2806; 3) the rename/allocator unit 2852 performs the allocation stage 2808 and renaming stage 2810; 4) the scheduler unit(s) 2856 performs the schedule stage 2812; 5) the physical register file(s) unit(s) 2858 and the memory unit 2870 perform the register read/memory read stage 2814; the execution cluster 2860 perform the execute stage 2816; 6) the memory unit 2870 and the physical register file(s) unit(s) 2858 perform the write back/memory write stage 2818; 7) various units may be involved in the exception handling stage 2822; and 8) the retirement unit 2854 and the physical register file(s) unit(s) 2858 perform the commit stage 2824.
The core 2890 may support one or more instructions sets (e.g., the x86 instruction set (with some extensions that have been added with newer versions); the MIPS instruction set of MIPS Technologies of Sunnyvale, Calif.; the ARM instruction set (with optional additional extensions such as NEON) of ARM Holdings of Sunnyvale, Calif.), including the instruction(s) described herein. In one embodiment, the core 2890 includes logic to support a packed data instruction set extension (e.g., AVX1, AVX2), thereby allowing the operations used by many multimedia applications to be performed using packed data.
It should be understood that the core may support multithreading (executing two or more parallel sets of operations or threads), and may do so in a variety of ways including time sliced multithreading, simultaneous multithreading (where a single physical core provides a logical core for each of the threads that physical core is simultaneously multithreading), or a combination thereof (e.g., time sliced fetching and decoding and simultaneous multithreading thereafter such as in the Intel® Hyperthreading technology).
While register renaming is described in the context of out-of-order execution, it should be understood that register renaming may be used in an in-order architecture. While the illustrated embodiment of the processor also includes separate instruction and data cache units 2834/2874 and a shared L2 cache unit 2876, alternative embodiments may have a single internal cache for both instructions and data, such as, for example, a Level 1 (L1) internal cache, or multiple levels of internal cache. In some embodiments, the system may include a combination of an internal cache and an external cache that is external to the core and/or the processor. Alternatively, all of the cache may be external to the core and/or the processor.
Specific Exemplary in-Order Core Architecture
The local subset of the L2 cache 2904 is part of a global L2 cache that is divided into separate local subsets, one per processor core. Each processor core has a direct access path to its own local subset of the L2 cache 2904. Data read by a processor core is stored in its L2 cache subset 2904 and can be accessed quickly, in parallel with other processor cores accessing their own local L2 cache subsets. Data written by a processor core is stored in its own L2 cache subset 2904 and is flushed from other subsets, if necessary. The ring network ensures coherency for shared data. The ring network is bi-directional to allow agents such as processor cores, L2 caches and other logic blocks to communicate with each other within the chip. Each ring data-path is 1012-bits wide per direction.
Thus, different implementations of the processor 3000 may include: 1) a CPU with the special purpose logic 3008 being integrated graphics and/or scientific (throughput) logic (which may include one or more cores), and the cores 3002A-N being one or more general purpose cores (e.g., general purpose in-order cores, general purpose out-of-order cores, a combination of the two); 2) a coprocessor with the cores 3002A-N being a large number of special purpose cores intended primarily for graphics and/or scientific (throughput); and 3) a coprocessor with the cores 3002A-N being a large number of general purpose in-order cores. Thus, the processor 3000 may be a general-purpose processor, coprocessor or special-purpose processor, such as, for example, a network or communication processor, compression engine, graphics processor, GPGPU (general purpose graphics processing unit), a high-throughput many integrated core (MIC) coprocessor (including 30 or more cores), embedded processor, or the like. The processor may be implemented on one or more chips. The processor 3000 may be a part of and/or may be implemented on one or more substrates using any of a number of process technologies, such as, for example, BiCMOS, CMOS, or NMOS.
The memory hierarchy includes one or more levels of cache within the cores, a set of one or more shared cache units 3006, and external memory (not shown) coupled to the set of integrated memory controller units 3014. The set of shared cache units 3006 may include one or more mid-level caches, such as level 2 (L2), level 3 (L3), level 4 (L4), or other levels of cache, a last level cache (LLC), and/or combinations thereof. While in one embodiment a ring based interconnect unit 3012 interconnects the integrated graphics logic 3008 (integrated graphics logic 3008 is an example of and is also referred to herein as special purpose logic), the set of shared cache units 3006, and the system agent unit 3010/integrated memory controller unit(s) 3014, alternative embodiments may use any number of well-known techniques for interconnecting such units. In one embodiment, coherency is maintained between one or more cache units 3006 and cores 3002-A-N.
In some embodiments, one or more of the cores 3002A-N are capable of multithreading. The system agent 3010 includes those components coordinating and operating cores 3002A-N. The system agent unit 3010 may include for example a power control unit (PCU) and a display unit. The PCU may be or include logic and components needed for regulating the power state of the cores 3002A-N and the integrated graphics logic 3008. The display unit is for driving one or more externally connected displays.
The cores 3002A-N may be homogenous or heterogeneous in terms of architecture instruction set; that is, two or more of the cores 3002A-N may be capable of execution the same instruction set, while others may be capable of executing only a subset of that instruction set or a different instruction set.
Exemplary Computer Architectures
Referring now to
The optional nature of additional processors 3115 is denoted in
The memory 3140 may be, for example, dynamic random access memory (DRAM), phase change memory (PCM), or a combination of the two. For at least one embodiment, the controller hub 3120 communicates with the processor(s) 3110, 3115 via a multi-drop bus, such as a frontside bus (FSB), point-to-point interface such as QuickPath Interconnect (QPI), or similar connection 3195.
In one embodiment, the coprocessor 3145 is a special-purpose processor, such as, for example, a high-throughput MIC processor, a network or communication processor, compression engine, graphics processor, GPGPU, embedded processor, or the like. In one embodiment, controller hub 3120 may include an integrated graphics accelerator.
There can be a variety of differences between the physical resources 3110, 3115 in terms of a spectrum of metrics of merit including architectural, microarchitectural, thermal, power consumption characteristics, and the like.
In one embodiment, the processor 3110 executes instructions that control data processing operations of a general type. Embedded within the instructions may be coprocessor instructions. The processor 3110 recognizes these coprocessor instructions as being of a type that should be executed by the attached coprocessor 3145. Accordingly, the processor 3110 issues these coprocessor instructions (or control signals representing coprocessor instructions) on a coprocessor bus or other interconnect, to coprocessor 3145. Coprocessor(s) 3145 accept and execute the received coprocessor instructions.
Referring now to
Processors 3270 and 3280 are shown including integrated memory controller (IMC) units 3272 and 3282, respectively. Processor 3270 also includes as part of its bus controller units point-to-point (P-P) interfaces 3276 and 3278; similarly, second processor 3280 includes P-P interfaces 3286 and 3288. Processors 3270, 3280 may exchange information via a point-to-point (P-P) interface 3250 using P-P interface circuits 3278, 3288. As shown in
Processors 3270, 3280 may each exchange information with a chipset 3290 via individual P-P interfaces 3252, 3254 using point to point interface circuits 3276, 3294, 3286, 3298. Chipset 3290 may optionally exchange information with the coprocessor 3238 via a high-performance interface 3292. In one embodiment, the coprocessor 3238 is a special-purpose processor, such as, for example, a high-throughput MIC processor, a network or communication processor, compression engine, graphics processor, GPGPU, embedded processor, or the like.
A shared cache (not shown) may be included in either processor or outside of both processors yet connected with the processors via P-P interconnect, such that either or both processors' local cache information may be stored in the shared cache if a processor is placed into a low power mode.
Chipset 3290 may be coupled to a first bus 3216 via an interface 3296. In one embodiment, first bus 3216 may be a Peripheral Component Interconnect (PCI) bus, or a bus such as a PCI Express bus or another third generation I/O interconnect bus, although the scope of the present invention is not so limited.
As shown in
Referring now to
Referring now to
Embodiments of the mechanisms disclosed herein may be implemented in hardware, software, firmware, or a combination of such implementation approaches. Embodiments of the invention may be implemented as computer programs or program code executing on programmable systems comprising at least one processor, a storage system (including volatile and non-volatile memory and/or storage elements), at least one input device, and at least one output device.
Program code, such as code 3230 illustrated in
The program code may be implemented in a high level procedural or object oriented programming language to communicate with a processing system. The program code may also be implemented in assembly or machine language, if desired. In fact, the mechanisms described herein are not limited in scope to any particular programming language. In any case, the language may be a compiled or interpreted language.
One or more aspects of at least one embodiment may be implemented by representative instructions stored on a machine-readable medium which represents various logic within the processor, which when read by a machine causes the machine to fabricate logic to perform the techniques described herein. Such representations, known as “IP cores” may be stored on a tangible, machine readable medium and supplied to various customers or manufacturing facilities to load into the fabrication machines that actually make the logic or processor.
Such machine-readable storage media may include, without limitation, non-transitory, tangible arrangements of articles manufactured or formed by a machine or device, including storage media such as hard disks, any other type of disk including floppy disks, optical disks, compact disk read-only memories (CD-ROMs), compact disk rewritable's (CD-RWs), and magneto-optical disks, semiconductor devices such as read-only memories (ROMs), random access memories (RAMs) such as dynamic random access memories (DRAMs), static random access memories (SRAMs), erasable programmable read-only memories (EPROMs), flash memories, electrically erasable programmable read-only memories (EEPROMs), phase change memory (PCM), magnetic or optical cards, or any other type of media suitable for storing electronic instructions.
Accordingly, embodiments of the invention also include non-transitory, tangible machine-readable media containing instructions or containing design data, such as Hardware Description Language (HDL), which defines structures, circuits, apparatuses, processors and/or system features described herein. Such embodiments may also be referred to as program products.
Emulation (Including Binary Translation, Code Morphing, Etc.)
In some cases, an instruction converter may be used to convert an instruction from a source instruction set to a target instruction set. For example, the instruction converter may translate (e.g., using static binary translation, dynamic binary translation including dynamic compilation), morph, emulate, or otherwise convert an instruction to one or more other instructions to be processed by the core. The instruction converter may be implemented in software, hardware, firmware, or a combination thereof. The instruction converter may be on processor, off processor, or part on and part off processor.
Example 1 provides an exemplary single instruction, multiple data (SIMD) processor comprising: fetch circuitry to fetch one or more instructions until a set of instructions has been fetched, each instruction to specify source and destination tile operands, decode circuitry to decode the set of fetched instructions and execution circuitry, responsive to the set of decoded instructions to: identify first and second decoded instructions belonging to a chain of instructions, dynamically select and configure a SIMD path comprising first and second processing engines (PE) to execute the first and second decoded instructions and set aside the specified destination of the first decoded instruction, and instead route a result of the first decoded instruction from the first PE to be used by the second PE to perform the second decoded instruction.
Example 2 includes the substance of the exemplary processor of Example 1, wherein the set of instructions comprises at least one instruction, and wherein each of the chain of instructions is further to comprise a chain control field to cause the execution circuitry to determine that the instruction is part of a chain and wherein each of the chain control fields is to indicate a chain position hint comprising one of a start of chain hint, a middle of chain hint, and an end of chain hint to mark a last instruction in the chain, wherein the chain control fields are to cause the execution circuitry to determine existence of the chain, to dynamically select and configure the SIMD path to perform the chain of instructions, to set aside the specified destinations of all but the last instruction in the chain, and to route results of all but the last instruction to the PE performing a next instruction in the chain.
Example 3 includes the substance of the exemplary processor of Example 1, wherein the first and second decoded instructions specify different opcodes to specify different, first and second arithmetic operations, and wherein the selected first and second processing engines are to have limited functionality, specializing in the first arithmetic operation and the second arithmetic operation, respectively.
Example 4 includes the substance of the exemplary processor of Example 2, wherein a first instruction of the chain of instructions is further to comprise a chain control field, the chain control field comprising a chain header to identify the first instruction and one or more subsequent instructions as being part of the chain or instructions.
Example 5 includes the substance of the exemplary processor of Example 1, wherein the specified source and destination matrices (tiles) each comprises M rows and N columns of elements and is only to contain valid elements in one row.
Example 6 includes the substance of the exemplary processor of Example 1, wherein the execution circuitry is further to maintain a record of instructions having specified destinations set aside, and to roll back and re-execute a set of recorded instructions as needed.
Example 7 includes the substance of the exemplary processor of Example 1, wherein the execution circuitry is further to save a machine state before the setting aside, and to subsequently execute a machine state consistency check to detect any inconsistencies caused by the setting aside.
Example 8 includes the substance of the exemplary processor of Example 1, wherein the set of instructions comprises at least two instructions, the execution circuitry further comprising a binary translator to buffer the set of instructions and to detect data dependencies among two or more instructions to determine that the two or more instructions are part of the chain.
Example 9 includes the substance of the exemplary processor of Example 1, wherein the execution circuitry is further to avoid alerting an operating system to the setting aside.
Example 10 includes the substance of the exemplary processor of Example 1, wherein the tile specified as the destination of the first instructions further comprises a tile configuration, and wherein the execution circuitry is further to set a dirty bit in the tile configuration before the setting aside, and to cause a fault if the dirty tile is read by a subsequent instruction.
Example 11 provides an exemplary method performed by a single instruction, multiple data (SIMD) processor, comprising: fetching, using fetch circuitry, one or more instructions comprising a set of instructions, each specifying source and destination matrices (tiles), decoding, using decode circuitry, the set of fetched instructions and executing, using execution circuitry, the set of decoded instructions, to: identify first and second decoded instructions belonging to a chain of instructions, dynamically select and configure a SIMD path comprising first and second processing engines (PE) to execute the first and second decoded instructions and set aside the specified destination of the first decoded instruction, and instead route a result of the first decoded instruction from the first PE to be used by the second PE to perform the second decoded instruction.
Example 12 includes the substance of the exemplary method of Example 11, wherein the set of instructions comprises at least one instruction, and wherein each of the chain of instructions is further to comprise a chain control field to cause the execution circuitry to determine that that the instruction is part of a chain and wherein each of the chain control fields is to indicate a chain position hint comprising one of a start of chain hint, a middle of chain hint, and an end of chain hint to mark a last instruction in the chain, wherein the chain control fields are to cause the execution circuitry to determine existence of the chain, to dynamically select and configure the SIMD path to perform the chain of instructions, to set aside the specified destinations of all but the last instruction in the chain, and to route results of all but the last instruction to the PE performing a next instruction in the chain.
Example 13 includes the substance of the exemplary method of Example 11, wherein the first and second decoded instructions specify different opcodes to specify different, first and second arithmetic operations, and wherein the selected first and second processing engines are to have limited functionality, specializing in the first arithmetic operation and the second arithmetic operation, respectively.
Example 14 includes the substance of the exemplary method of Example 11, wherein a first instruction of the chain of instructions is further to comprise a chain control field, the chain control field comprising a chain header to identify the first instruction and one or more subsequent instructions as being part of the chain or instructions.
Example 15 includes the substance of the exemplary method of Example 11, wherein the specified source and destination matrices (tiles) each comprises M rows and N columns of elements and is only to contain valid elements in one row.
Example 16 includes the substance of the exemplary method of Example 11, wherein the execution circuitry is further to maintain a record of instructions having specified destinations set aside, and to roll back and re-execute one or more of the recorded instructions as needed.
Example 17 includes the substance of the exemplary method of Example 11, wherein the execution circuitry is further to save a machine state before the setting aside, and to subsequently execute a machine state consistency check to detect any inconsistencies caused by the setting aside.
Example 18 includes the substance of the exemplary method of Example 11, wherein the set of instructions comprises at least two instructions, the execution circuitry further comprising a binary translator to buffer the set of instructions and to detect data dependencies among two or more instructions to determine that the two or more instructions are part of the chain.
Example 19 includes the substance of the exemplary method of Example 11, wherein the execution circuitry is further to avoid alerting an operating system to the setting aside.
Example 20 includes the substance of the exemplary method of Example 11, wherein the tile specified as the destination of the first instructions further comprises a tile configuration, and wherein the execution circuitry is further to set a dirty bit in the tile configuration before the setting aside, and to cause a fault if the dirty tile is read by a subsequent instruction.
Example 21 provides an exemplary apparatus comprising: means for fetching one or more instructions until a set of instructions has been fetched, each instruction to specify source and destination tile operands, means for decoding the set of fetched instructions and means for executing, responsive to the set of decoded instructions to: identify first and second decoded instructions belonging to a chain of instructions, dynamically select and configure a SIMD path comprising first and second processing engines (PE) to execute the first and second decoded instructions and set aside the specified destination of the first decoded instruction, and instead route a result of the first decoded instruction from the first PE to be used by the second PE to perform the second decoded instruction.
Example 22 includes the substance of the exemplary apparatus of Example 21, wherein the set of instructions comprises at least one instruction, and wherein each of the chain of instructions is further to comprise a chain control field to cause the means for executing to determine that that the instruction is part of a chain and wherein each of the chain control fields is to indicate a chain position hint comprising one of a start of chain hint, a middle of chain hint, and an end of chain hint to mark a last instruction in the chain, wherein the chain control fields are to cause the means for executing to determine existence of the chain, to dynamically select and configure the SIMD path to perform the chain of instructions, to set aside the specified destinations of all but the last instruction in the chain, and to route results of all but the last instruction to the PE performing a next instruction in the chain.
Example 23 includes the substance of the exemplary apparatus of Example 21, wherein the first and second decoded instructions specify different opcodes to specify different, first and second arithmetic operations, and wherein the selected first and second processing engines are to have limited functionality, specializing in the first arithmetic operation and the second arithmetic operation, respectively.
Example 24 includes the substance of the exemplary apparatus of Example 22, wherein a first instruction of the chain of instructions is further to comprise a chain control field, the chain control field comprising a chain header to identify the first instruction and one or more subsequent instructions as being part of the chain or instructions.
Example 25 includes the substance of the exemplary apparatus of Example 21, wherein the specified source and destination matrices (tiles) each comprises M rows and N columns of elements and is only to contain valid elements in one row.
Example 26 includes the substance of the exemplary apparatus of Example 21, wherein the means for executing is further to maintain a record of instructions having specified destinations set aside, and to roll back and re-execute a set of recorded instructions as needed.
Example 27 includes the substance of the exemplary apparatus of Example 21, wherein the means for executing is further to save a machine state before the setting aside, and to subsequently execute a machine state consistency check to detect any inconsistencies caused by the setting aside.
Example 28 includes the substance of the exemplary apparatus of Example 21, wherein the set of instructions comprises at least two instructions, the means for executing further comprising a binary translator to buffer the set of instructions and to detect data dependencies among two or more instructions to determine that the two or more instructions are part of the chain.
Example 29 includes the substance of the exemplary apparatus of Example 21, wherein the means for executing is further to avoid alerting an operating system to the setting aside.
Example 30 includes the substance of the exemplary apparatus of Example 21, wherein the tile specified as the destination of the first instructions further comprises a tile configuration, and wherein the means for executing is further to set a dirty bit in the tile configuration before the setting aside, and to cause a fault if the dirty tile is read by a subsequent instruction.
Example 31 provides an exemplary non-transitory machine-readable medium containing instructions, when accessed by a single instruction, multiple data (SIMD) processor, cause the processor to respond by: fetching, using fetch circuitry, one or more instructions comprising a set of instructions, each specifying source and destination matrices (tiles), decoding, using decode circuitry, the set of fetched instructions and executing, using execution circuitry, the set of decoded instructions, to: identify first and second decoded instructions belonging to a chain of instructions, dynamically select and configure a SIMD path comprising first and second processing engines (PE) to execute the first and second decoded instructions and set aside the specified destination of the first decoded instruction, and instead route a result of the first decoded instruction from the first PE to be used by the second PE to perform the second decoded instruction.
Example 32 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the set of instructions comprises at least one instruction, and wherein each of the chain of instructions is further to comprise a chain control field to cause the execution circuitry to determine that that the instruction is part of a chain and wherein each of the chain control fields is to indicate a chain position hint comprising one of a start of chain hint, a middle of chain hint, and an end of chain hint to mark a last instruction in the chain, wherein the chain control fields are to cause the execution circuitry to determine existence of the chain, to dynamically select and configure the SIMD path to perform the chain of instructions, to set aside the specified destinations of all but the last instruction in the chain, and to route results of all but the last instruction to the PE performing a next instruction in the chain.
Example 33 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the first and second decoded instructions specify different opcodes to specify different, first and second arithmetic operations, and wherein the selected first and second processing engines are to have limited functionality, specializing in the first arithmetic operation and the second arithmetic operation, respectively.
Example 34 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein a first instruction of the chain of instructions is further to comprise a chain control field, the chain control field comprising a chain header to identify the first instruction and one or more subsequent instructions as being part of the chain or instructions.
Example 35 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the specified source and destination matrices (tiles) each comprises M rows and N columns of elements and is only to contain valid elements in one row.
Example 36 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the processor is to further respond by maintaining a record of instructions having specified destinations set aside and rolling back and re-executing one or more of the recorded instructions as needed.
Example 37 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the processor is to further respond by saving a machine state before the setting aside, and subsequently executing a machine state consistency check to detect any inconsistencies caused by the setting aside.
Example 38 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the set of instructions comprises at least two instructions, and the processor is to further respond by using a binary translator to buffer the set of instructions and to detect data dependencies among two or more instructions to determine that the two or more instructions are part of the chain.
Example 39 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the processor is to further respond by avoiding alerting an operating system to the setting aside.
Example 40 includes the substance of the exemplary non-transitory machine-readable medium of Example 31, wherein the tile specified as the destination of the first instructions further comprises a tile configuration, and wherein the processor is to further respond by setting a dirty bit in the tile configuration before the setting aside and causing a fault if the dirty tile is read by a subsequent instruction.
| Number | Name | Date | Kind |
|---|---|---|---|
| 6704925 | Bugnion | Mar 2004 | B1 |
| 20020069346 | Farrell | Jun 2002 | A1 |
| 20020108026 | Balmer | Aug 2002 | A1 |
| 20060242388 | Tremblay | Oct 2006 | A1 |
| 20090210669 | Luick | Aug 2009 | A1 |
| 20130246712 | Liu | Sep 2013 | A1 |
| 20140281386 | Fox et al. | Sep 2014 | A1 |
| 20150220346 | Wang et al. | Aug 2015 | A1 |
| 20160179551 | Barowski et al. | Jun 2016 | A1 |
| 20170083314 | Burger | Mar 2017 | A1 |
| Entry |
|---|
| “Wolfram Language & System” p. 2, https://reference.wolfram.com/language/ref/Transpose.html (2017) (Year: 2017). |
| Anonymous: “Tile processor—Wikipedia”, Aug. 30, 2017 (Aug. 30, 2017), XP055607501, Retrieved from the Internet: URL:https://en.wikipedia.org/w/index.php?title= Tile processor&oldid= 798055608 [retrieved on Jul. 22, 2019]. |
| European Search Report and Search Opinion Received for EP Application No. 19157043.1, dated Aug. 12, 2019, 12 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20190303167 A1 | Oct 2019 | US |
