Patent Grant
9665490
This application is related to the following co-pending U.S. patent applications, each of which has a common assignee and common inventors.
Field of the Invention
This invention relates in general to the field of microelectronics, and more particularly to apparatus and methods for restoring compressed cache repair data following a multi-core power gating event.
Description of the Related Art
Integrated device technologies have exponentially advanced over the past 40 years. More specifically directed to the microprocessor fields, starting with 4-bit, single instruction, 10-micrometer devices, the advances in semiconductor fabrication technologies have enabled designers to provide increasingly more complex devices in terms of architecture and density. In the 80's and 90's, so-called pipeline microprocessors and superscalar microprocessors were developed comprising millions of transistors on a single die. And now 20 years later, 64-bit, 32-nanometer devices are being produced that have billions of transistors on a single die, and which comprise multiple microprocessor cores for the processing of data.
One requirement that has persisted since these early microprocessors were produced is the need to initialize these devices with configuration data when they are powered up or when they are reset. For example, many architectures enable devices to be configured to execute at one of many selectable frequencies and/or voltages. Other architectures require that each device have a serial number and other information that can be read via execution of an instruction. Yet other devices require initialization data for internal registers and control circuits. Still other microprocessors, particularly that with on-board cache memories, utilize repair data to implement redundant circuits within these memories to correct for fabrication errors.
As one skilled in the art will appreciate, designers have traditionally employed semiconductor fuse arrays on-die to store and provide initial configuration and repair data. These fuse arrays are generally programmed by blowing selected fuses therein after a part has been fabricated, and the arrays contain thousands of bits of information which are read by a corresponding device upon power-up/reset to initialize and configure the device for operation.
As device complexity has increased over the past years, the amount of configuration/repair data that is required for a typical device has proportionately increased. Yet, as one skilled in the art will appreciate, though transistor size shrinks in proportion to the semiconductor fabrication process employed, semiconductor fuse size increases due to the unique requirements for programming fuses on die. This phenomenon, in and of itself, is a problem for designers, who are prevalently constrained by real estate and power considerations. In other words, there is just not enough real estate on a given die to fabricate a huge fuse array.
In addition, the ability to fabricate multiple device cores on a single die has geometrically exacerbated the problem, because configuration requirements for each of the cores result in a requirement for a number of fuses on die, in a single array or distinct arrays, that is proportional to the number of cores disposed thereon.
Furthermore, as one skilled in the art will appreciate, multiple-core devices utilize complex power-saving modes of operation that result in one or more of the cores being powered down in a so-called power gating event (or, “sleep mode”) when not in use. Consequently, when a core is powered up following a power gating event, the same requirement for initialization, configuration, and repair persists, except that initialization speed requirements are much more stringent.
Therefore, what is needed is apparatus and methods that enable configuration/repair data to be stored and provided to a multi-core device that require significantly less real estate and power on a single die than that which has heretofore been provided.
In addition, what is needed is a fuse array mechanism that can store and provide significantly more configuration/repair data than current techniques while requiring the same or less real estate on a multi-core die.
Furthermore, what is needed is a technique that facilitates prompt initialization, configuration, and repair of a multi-core device following a power gating event.
The present invention provides a superior technique for restoring configuration data following a power gating event. In one aspect, an apparatus for providing configuration data to an integrated circuit is considered. The apparatus includes a semiconductor fuse array, a stores, and a plurality of cores. The semiconductor fuse array is disposed on a die, and is programmed with compressed configuration data. The stores, also disposed on the die, is configured for storage and access of decompressed configuration data sets. The plurality of cores is also disposed on the die, where each of the plurality of cores is coupled to the semiconductor fuse array, and where one of the plurality of cores is configured to access the semiconductor fuse array upon power-up/reset to read and decompress the compressed configuration data, and to store the decompressed configuration data sets for one or more cache memories within the each of the plurality of cores in the stores. Each of the plurality of cores includes reset logic and sleep logic. The reset logic is configured to employ the decompressed configuration data sets to initialize the one or more cache memories upon power-up/reset. The sleep logic is configured to determine that power is restored following a power gating event, and is configured to subsequently access the stores to retrieve and employ the decompressed configuration data sets to initialize the one or more caches following the power gating event.
One aspect of the present invention contemplates an apparatus for providing configuration data to an integrated circuit device. The apparatus includes a multi-core microprocessor. The multi-core microprocessor includes a semiconductor fuse array, a stores, and a plurality of cores. The semiconductor fuse array is disposed on a die, and is programmed with compressed configuration data. The stores, also disposed on the die, is configured for storage and access of decompressed configuration data sets. The plurality of cores is also disposed on the die, where each of the plurality of cores is coupled to the semiconductor fuse array, and where one of the plurality of cores is configured to access the semiconductor fuse array upon power-up/reset to read and decompress the compressed configuration data, and to store the decompressed configuration data sets for one or more cache memories within the each of the plurality of cores in the stores. Each of the plurality of cores includes reset logic and sleep logic. The reset logic is configured to employ the decompressed configuration data sets to initialize the one or more cache memories upon power-up/reset. The sleep logic is configured to determine that power is restored following a power gating event, and is configured to subsequently access the stores to retrieve and employ the decompressed configuration data sets to initialize the one or more caches following the power gating event.
Another aspect of the present invention comprehends a method for configuring an integrated circuit. The method includes first disposing a semiconductor fuse array on a die, into which is programmed compressed configuration data; second disposing a stores on the die, configured for storage and access of decompressed configuration data sets; third disposing a plurality of microprocessor cores on the die, where each of the plurality of microprocessor cores is coupled to the semiconductor fuse array, and where one of the plurality of microprocessor cores is configured to access the semiconductor fuse array upon power-up/reset to read and decompress the compressed configuration data, and to store the decompressed configuration data sets for one or more cache memories within the each of the plurality of cores in the stores; via reset logic disposed within the each of the plurality of cores, employing the decompressed configuration data sets to initialize the one or more cache memories upon power-up/reset; and via sleep logic disposed within the each of the plurality of cores, determining that power is restored following a power gating event, and subsequently accessing the stores to retrieve and employ the decompressed configuration data sets to initialize the one or more caches following the power gating event.
Regarding industrial applicability, the present invention is implemented within a MICROPROCESSOR which may be used in a general purpose or special purpose computing device.
These and other objects, features, and advantages of the present invention will become better understood with regard to the following description, and accompanying drawings where:
Exemplary and illustrative embodiments of the invention are described below. In the interest of clarity, not all features of an actual implementation are described in this specification, for those skilled in the art will appreciate that in the development of any such actual embodiment, numerous implementation specific decisions are made to achieve specific goals, such as compliance with system-related and business related constraints, which vary from one implementation to the next. In addition, it will be appreciated that such a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure. Various modifications to the preferred embodiment will be apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments. Therefore, the present invention is not intended to be limited to the particular embodiments shown and described herein, but is to be accorded the widest scope consistent with the principles and novel features herein disclosed.
The present invention will now be described with reference to the attached figures. Various structures, systems, and devices are schematically depicted in the drawings for purposes of explanation only and so as to not obscure the present invention with details that are well known to those skilled in the art. Nevertheless, the attached drawings are included to describe and explain illustrative examples of the present invention. The words and phrases used herein should be understood and interpreted to have a meaning consistent with the understanding of those words and phrases by those skilled in the relevant art. No special definition of a term or phrase (i.e., a definition that is different from the ordinary and customary meaning as understood by those skilled in the art) is intended to be implied by consistent usage of the term or phrase herein. To the extent that a term or phrase is intended to have a special meaning (i.e., a meaning other than that understood by skilled artisans) such a special definition will be expressly set forth in the specification in a definitional manner that directly and unequivocally provides the special definition for the term or phrase.
In view of the above background discussion on device fuse arrays and associated techniques employed within present day integrated circuits for providing configuration data during initial power-up, a discussion of the limitations and disadvantages of those techniques will be presented with reference to
Integrated Circuit (IC): A set of electronic circuits fabricated on a small piece of semiconductor material, typically silicon. An IC is also referred to as a chip, a microchip, or a die.
Central Processing Unit (CPU): The electronic circuits (i.e., “hardware”) that execute the instructions of a computer program (also known as a “computer application,” “application program,” “program,” or “application”) by performing operations on data that include arithmetic operations, logical operations, and input/output operations.
Microprocessor: An electronic device that functions as a CPU on a single integrated circuit. A microprocessor receives digital data as input, processes the data according to instructions fetched from a memory (either on-die or off-die), and generates results of operations prescribed by the instructions as output. A general purpose microprocessor may be employed in devices including, but not limited to, a desktop, mobile, or tablet computer, and may be utilized for tasks such as, but not limited to, computation, text editing, multimedia display, and Internet browsing. A microprocessor may also be disposed in an embedded system to control a wide variety of devices that include appliances, mobile telephones, smart phones, and industrial control devices.
Multi-Core Processor: Also known as a multi-core microprocessor, a multi-core processor is a microprocessor having multiple CPUs (also known as “cores”) fabricated on a single integrated circuit.
Instruction Set Architecture (ISA) or Instruction Set: A part of a computer architecture related to programming that includes data types, instructions, registers, addressing modes, memory architecture, interrupt and exception handling, and input/output. An ISA includes a specification of the set of opcodes (i.e., machine language instructions), and the native commands implemented by a particular CPU.
x86-Compatible Microprocessor: A microprocessor capable of executing computer applications that are programmed according to the x86 ISA.
Microcode: A term employed to refer to a plurality of micro instructions. A micro instruction (also referred to as a “native instruction”) is an instruction at the level that a microprocessor sub-unit executes. Exemplary sub-units include integer units, floating point units, MMX units, and load/store units. For example, micro instructions are directly executed by a reduced instruction set computer (RISC) microprocessor. For a complex instruction set computer (CISC) microprocessor such as an x86-compatible microprocessor, x86 instructions are translated into associated micro instructions, and the associated micro instructions are directly executed by a sub-unit or sub-units within the CISC microprocessor.
Fuse: A conductive structure typically arranged as a filament which can be broken at select locations by applying a voltage across the filament and/or current through the filament. Fuses may be deposited at specified areas across a die topography using well known fabrication techniques to produce filaments at all potential programmable areas. A fuse structure is blown (or unblown) subsequent to fabrication to provide for desired programmability of a corresponding device disposed on the die.
Turning to
As one skilled in the art will appreciate, fuses (also called “links” or “fuse structures”) are employed in a vast number of present day integrated circuit devices to provide for configuration of the devices after the devices have been fabricated. For example, consider that the microprocessor core 101 of
The above example is merely one of many different uses for configuration fuses in an integrated circuit device such as a microprocessor core 101 of
Fuse arrays 102 provide an excellent means for configuring a device such as a microprocessor core 101 subsequent to fabrication of the device. By blowing selected fuses in the fuse array 102, the core 101 can be configured for operation in its intended environment. Yet, as one skilled in the art will appreciate, operating environments may change following programming of the fuse array 102. Business requirements may dictate that a device 101 originally configured as, say, as desktop device 101, be reconfigured as a mobile device 101. Accordingly, designers have provided techniques that utilize redundant banks of fuses within the fuse array 102 to provide for “unblowing” selected fuses therein, thus enabling the device 101 to be reconfigured, fabrication errors to be corrected, and etc. These redundant array techniques will now be discussed with reference to
Referring now to
The fuse array 201 is coupled to a set of registers 210-211 that are typically disposed within reset logic in the core 101. A primary register PR1 is employed to read one of the first fuse banks PFB1-PFBN (say, PFB3 as is shown in the diagram 200) and a redundant register RR1 is employed to read a corresponding one of the redundant fuse banks RFB1-RFBN. The registers 210-211 are coupled to exclusive-OR logic 212 that generates an output FB3.
In operation, subsequent to fabrication of the core 101, the first fuse banks PFB1-PFBN are programmed by known techniques with configuration data for the core 101. The redundant fuse banks RFB1-RFBN are not blown and remain at a logic low state for all fuses therein. Upon power-up/reset of the core 101, both the first fuse banks PFB1-PFBN and the redundant fuse banks RFB1-RFBN are read as required for configuration into the primary and redundant registers 210-211, respectively. The exclusive-OR logic 212 generates the output FB3 that is a logical exclusive-OR result of the contents of the registers 210-211. Since all of the redundant fuse banks are unblown (i.e., logic low states), the output FB3 value is simply that which was programmed into the first fuse banks PFB1-PFBN subsequent to fabrication.
Consider now, though, that design or business requirements dictate that some of the information that was programmed into the first fuse banks PFB1-PFBN needs to change. Accordingly, a programming operation is performed to blow corresponding fuses 203 within the redundant fuse banks RFB1-RFBN in order to change the information that is read at power-up. By blowing a fuse 203 in a selected redundant bank RFB1-RFBN, the value of a corresponding fuse 203 in the primary fuse bank PFB1-PFBN is logically complemented.
The mechanism of
Heretofore, the fuse array mechanisms as discussed above with reference to
Both of the above limitations, and others, pose significant challenges to device designers, and more specifically to multi-core device designers, and the present inventors note that significant improvements over conventional device configuration mechanisms can be implemented in accordance with the present invention, which allows for programming of individual cores in a multi-core device along with substantial increases in cache correction and fuse reprogramming (“reblow”) elements. The present invention will now be discussed with reference to
Turning to
In one embodiment, the cores 332 comprise microprocessor cores configured as a multi-core microprocessor 330. In another embodiment, the multi-core microprocessor 330 is configured as an x86-compatible multi-core microprocessor. In yet another embodiment, the cache 334 comprises a level 2 (L2) cache 334 associated with the microprocessor cores 332. In one embodiment, the fuse array 336 comprises 8192 (8K) individual fuses (not shown), although other numbers of fuses are contemplated. In a single-core embodiment, only one core 332 is disposed on the die 330 and the core 332 is coupled to the cache 334 and physical fuse array 336. The present inventors note that although features and functions of the present invention will henceforth be discussed in the context of a multi-core device 330, these features and functions are equally applicable to a single-core embodiment as well.
The system 300 also includes a device programmer 310 that includes a compressor 320 that is coupled to a virtual fuse array 303. In one embodiment, the device programmer 310 may comprise a CPU (not shown) that is configured to process configuration data and to program the fuse array 336 following fabrication of the die 330 according to well known programming techniques. The CPU may be integrated into a wafer test apparatus that is employed to test the device die 330 following fabrication. In one embodiment, the compressor 320 may comprise an application program that executes on the device programmer 310 and the virtual fuse array 303 may comprise locations within a memory that is accessed by the compressor 320. The virtual fuse array 303 includes a plurality of virtual fuse banks 301, that each comprise a plurality of virtual fuses 302. In one embodiment the virtually fuse array 303 comprises 128 virtual fuse banks 301 that each comprise 64 virtual fuses 302, resulting in a virtual array 303 that is 8 Kb in size.
Operationally, configuration information for the device 330 is entered into the virtual fuse array 303 as part of the fabrication process, and as is described above with reference to
After the information is entered into the virtual fuse array 303, the compressor 320 reads the state of the virtual fuses 302 in each of the virtual fuse banks 301 and compresses the information using distinct compression algorithms corresponding to each of the data types to render compressed fuse array data 303. In one embodiment, system data for control circuits is not compressed, but rather is transferred without compression. To compress microcode register data, a microcode register data compression algorithm is employed that is effective for compressing data having a state distribution that corresponds to the microcode register data. To compress microcode patch data, a microcode patch data compression algorithm is employed that is effective for compressing data having a state distribution that corresponds to the microcode patch data. To compress cache correction data, a cache correction data compression algorithm is employed that is effective for compressing data having a state distribution that corresponds to the cache correction data.
The device programmer 310 then programs the uncompressed and compressed fuse array data into the physical fuse array 336 on the die 330.
Upon power-up/reset, each of the cores 332 may access the physical fuse array 336 to retrieve the uncompressed and compressed fuse array data, and reset circuits/microcode (not shown) disposed within each of the cores 332 distributes the uncompressed fuse array data, and decompresses the compressed fuse array data according to distinct decompression algorithms corresponding to each of the data types noted above to render values originally entered into the virtual fuse array 303. The reset circuits/microcode then enter the configuration information into control circuits (not shown), microcode registers (not shown), patch elements (not shown), and cache correction elements (not shown).
Advantageously, the fuse array compression system 300 according to the present invention enables device designers to employ substantially fewer numbers of fuses in a physical fuse array 336 over that which has heretofore been provided, and to utilize the compressed information programmed therein to configure a multi-core device 330 during power-up/reset.
Turning now to
The microprocessor core 420 comprises a reset controller 417 that receives a reset signal RESET which is asserted upon power-up of the core 420 and in response to events that cause the core 420 to initiate a reset sequence of steps. The reset controller 417 includes a decompressor 421. The decompressor 421 has a patch fuses element 408, a register fuses element 409, and a cache fuses element 410. The decompressor also comprises a fuse correction element 411 that is coupled to the patch fuses element 408, the register fuses element 409, and the cache fuses element 410 via bus 412. The patch fuses decompressor is coupled to microcode patch elements 414 in the core 420. The register fuses element 409 is coupled to microcode registers 415 in the core 420. And the cache fuses element 410 is coupled to cache correction elements 416 in the core 420. In one embodiment, the cache correction elements 416 are disposed within an on-die L2 cache (not shown) that is shared by all the cores 420, such as the cache 334 of
In operation, upon assertion of RESET the reset controller 416 reads the states of the fuses 402-406 in the physical fuse array 401 and distributes the states of the compressed system fuses 402 to the decompressor 421. After the fuse data has been read and distributed, the fuse correction element 411 of the decompressor 421 decompresses the compressed fuse correction fuses states to render data that indicates one or more fuse addresses in the physical fuse array 401 whose states are to be changed from that which was previously programmed. The data may also include a value for each of the one or more fuse addresses. The one or more fuse addresses (and optional values) are routed via bus 412 to the elements 408-410 so that the states of corresponding fuses processed therein are changed prior to decompression of their corresponding compressed data.
In one embodiment, the patch fuses element 408 comprises microcode that operates to decompress the states of the compressed microcode patch fuses 403 according to a microcode patch decompression algorithm that corresponds the microcode patch compression algorithm described above with reference to
In one embodiment, the reset controller 417 initiates execution of microcode within the patch fuses element 408 to decompress the states of the compressed microcode patch fuses 403. The reset controller 417 also initiates execution of microcode within the register fuses element 409 to decompress the states of the compressed register fuses 404. And the reset controller 417 further initiates execution of microcode within the cache fuses element 410 to decompress the states of the compressed cache correction fuses 406. The microcode within the decompressor 421 also operates to change the states of any fuses addressed by fuse correction data provided by the compressed fuse correction fuses 406 prior to decompression of the compressed data.
The reset controller 417, decompressor 421, and elements 408-411 therein according to the present invention are configured to perform the functions and operations as discussed above. The reset controller 417, decompressor 421, and elements 408-411 therein may comprise logic, circuits, devices, or microcode, or a combination of logic, circuits, devices, or microcode, or equivalent elements that are employed to execute the functions and operations according to the present invention as noted. The elements employed to accomplish these operations and functions within the reset controller 417, decompressor 421, and elements 408-411 therein may be shared with other circuits, microcode, etc., that are employed to perform other functions and/or operations within the reset controller 417, decompressor 421, and elements 408-411 therein or with other elements within the core 420.
After the states of the fuses 403-406 within the physical fuse array 401 have been changed and decompressed, the states of the decompressed “virtual” fuses are then routed, as appropriate to the microcode patch elements 414, the microcode registers 415, and the cache correction elements 416. Accordingly, the core 420 is configured for operation following completion of a reset sequence.
The present inventors note that the decompression functions discussed above need not necessarily be performed in a particular order during a reset sequence. For example, microcode patches may be decompressed following decompression of microcode registers initialization data. Likewise, the decompression functions may be performed in parallel or in an order suitable to satisfy design constraints.
Furthermore, the present inventors note that the implementations of the elements 408-411 need not necessarily be implemented in microcode versus hardware circuits, since in a typical microprocessor core 420 there exist elements of the core 420 which can more easily be initialized via hardware (such as a scan chain associated with a cache) as opposed to direct writes by microcode. Such implementation details are left up to designer judgment. However, the present inventors submit that the prior art teaches that cache correction fuses are conventionally read and entered into a cache correction scan chain by hardware circuits during reset prior to initiating the execution of microcode, and it is a feature of the present invention to implement the cache fuses decompressor 410 in microcode as opposed to hardware control circuits since a core's caches are generally not turned on until microcode runs. By utilizing microcode to implement the cache fuses element 410, a more flexible and advantageous mechanisms is provided for entering cache correction data into a scan chain, and significant hardware is saved.
Now referring to
The compressed configuration data 500 comprises one or more compressed data fields 502 for each of the configuration data types discussed above and are demarcated by end-of-type fields 503. Programming events (i.e., “blows”) are demarcated by an end-of-blow field 504. The compressed data fields 502 associated with each of the data types are encoded according to a compression algorithm that is optimized to minimize the number of bits (i.e., fuses) that are required to store the particular bit patterns associated with each of the data types. The number of fuses in the physical fuse array 336 that make up each of compressed data fields 502 is a function of the compression algorithm that is employed for a particular data type. For example, consider a core that comprises sixty-four 64-bit microcode registers which must be initialized to, say, all ones or all zeros. An optimum compression algorithm may be employed to yield 64 compressed data fields 502 for that data type, where each of the compressed data fields 502 comprises initialization data for a particular microcode register where the compressed data fields 502 are prescribed in register number order (i.e., 1-64). And each of the compressed data fields 502 comprises a single fuse which is blown if a corresponding microcode register is initialized to all ones, and which is not blown if the corresponding microcode register is initialized to all zeros.
The elements 408-410 of the decompressor 421 in the core 420 are configured to utilize the end-of-type fields 503 to determine where their respective compressed data is located within the physical fuse array 336 and the fuse correction decompressor 411 is configured to utilize the end-of-blow fields 504 to locate compressed fuse correction data that has been programmed (i.e., blown) subsequent to an initial programming event. It is a feature of the present invention to provide a substantial amount of spare fuses in the physical fuse array 336 to allow for a significant number of subsequent programming events, as will be discussed in more detail below.
The exemplary compressed type format discussed above is presented to clearly teach aspects of the present invention that are associated with compression and decompression of configuration data. However, the manner in which specific type data is compressed, demarcated, and the number and types of data to be compressed within the fuse array 401 is not intended to be restricted to the example of
Turning now to
Now turning to
Referring now to
Turning to
Turning now to
The present inventors have also observed that the real estate and power gains associated with utilization of a shared physical fuse array within which compressed configuration data is stored presents opportunities for additional features disposed on a multi-core die. In addition, the present inventors have noted that, as one skilled in the art will appreciate, present day semiconductor fuse structures often suffer from several shortcomings, one of which is referred to as “growback.” Growback is the reversal of the programming process such that a fuse will, after some time, reconnect after it has been blown, that is, it goes from a programmed (i.e., blown) state back to an unprogrammed (i.e., unblown) state.
In another aspect, as alluded to above, the present inventors have noted that there may exist challenges when power gating techniques are employed to minimize power consumption across a multi-core die, such as the die 330 of
Attention is now directed to
For purposes of illustration, only four cores 1101, a single physical fuse array 1110, and a single uncore stores 1130 are shown, however the present inventors note that the novel and inventive concepts according to the present invention can be extended to a plurality of cores 1101, fuse arrays 1110, and uncore stores 1130 of any number. In one embodiment, the uncore stores 1130 comprises a random access memory (RAM) that retains power during a power gating event. In one embodiment, the uncore stores 1130 comprises a 4 KB RAM, though other sizes are contemplated.
In operation, the power control 1120 is configured to perform power gating in order to remove and restore power to one or more of the cores 1101. During power-up/reset, the reset logic 1105 on each of the cores 1101 is configured to perform, among other operations, configuration of the cores 1101 as is described above. In addition, the reset logic 1105 is configured to read the configuration data register 1104 to determine if a core 1101 is a master core or a slave core. If the configuration data indicates that a core 1101 is a slave core, as part of the reset process, the slave core waits until the master core signals over the synchronization bus SYNC that decompressed cache repair data for each of the cores 1101 has been read from the fuse array 1110 and has been written into corresponding sub-stores 1131-1134 within the uncore stores 1130. Upon indication over SYNC that the corresponding sub-stores 1131-1134 have been written, each of the slave cores reads their respective decompressed repair data from the corresponding sub-stores 1131-1134 and proceeds to configure their respective on-core caches as is described above. If the configuration data indicates that a core 1101 is a master core, then as part of power-up/reset, the master core reads cache correction data for all of the cores 1101 from the fuse array 1110, decompresses the compressed correction data for all of the cores 1101, and writes the decompressed cache repair data to the sub-stores 1131-1134 corresponding to each of the cores 1101. The master core then signals to the other cores 1101 over SYNC that writing of the decompressed cache repair data is completed.
During a power gating event, the power control 1120 removes power from one or more of the cores 1101 to the extent that power is also removed from the core's on-core caches 1102. However, power is not removed from the uncore stores 1130, thus preserving the decompressed repair data for each of the cores 1101. The sleep logic 1106 is configured to determine when power is restored to a respective core 1101 following a power gating event, to directly read cache repair data for its on-core caches from its respective sub-stores 1131-1134, and to configure its respective repair data stores 1103 for correction of its on-core caches 1102, thus drastically reducing the time required to return to operations following the power gating event, while concurrently substantially increasing the lifetime of the fuse array 1110.
Portions of the present invention and corresponding detailed description are presented in terms of software, or algorithms and symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the ones by which those of ordinary skill in the art effectively convey the substance of their work to others of ordinary skill in the art. An algorithm, as the term is used here, and as it is used generally, is conceived to be a self-consistent sequence of steps leading to a desired result. The steps are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of optical, electrical, or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
It should be borne in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise, or as is apparent from the discussion, terms such as “processing” or “computing” or “calculating” or “determining” or “displaying” or the like, refer to the action and processes of a computer system, a microprocessor, a central processing unit, or similar electronic computing device, that manipulates and transforms data represented as physical, electronic quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage, transmission or display devices.
Note also that the software implemented aspects of the invention are typically encoded on some form of program storage medium or implemented over some type of transmission medium. The program storage medium may be electronic (e.g., read only memory, flash read only memory, electrically programmable read only memory), random access memory magnetic (e.g., a floppy disk or a hard drive) or optical (e.g., a compact disk read only memory, or “CD ROM”), and may be read only or random access. Similarly, the transmission medium may be metal traces, twisted wire pairs, coaxial cable, optical fiber, or some other suitable transmission medium known to the art. The invention is not limited by these aspects of any given implementation.
The particular embodiments disclosed above are illustrative only, and those skilled in the art will appreciate that they can readily use the disclosed conception and specific embodiments as a basis for designing or modifying other structures for carrying out the same purposes of the present invention, and that various changes, substitutions and alterations can be made herein without departing from the scope of the invention as set forth by the appended claims.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5155484 | Chambers, IV | Oct 1992 | A |
| 5513346 | Satagopan et al. | Apr 1996 | A |
| 5745734 | Craft et al. | Apr 1998 | A |
| 5784625 | Walker | Jul 1998 | A |
| 5889679 | Henry et al. | Mar 1999 | A |
| 6028445 | Lawman | Feb 2000 | A |
| 6330712 | Iwaya | Dec 2001 | B1 |
| 6349395 | Ohuchi et al. | Feb 2002 | B2 |
| 6525678 | Veenstra et al. | Feb 2003 | B1 |
| 6577156 | Anand et al. | Jun 2003 | B2 |
| 6748518 | Guthrie et al. | Jun 2004 | B1 |
| 6772356 | Qureshi et al. | Aug 2004 | B1 |
| 6888774 | Suzuki et al. | May 2005 | B2 |
| 6924663 | Masui et al. | Aug 2005 | B2 |
| 7251756 | Anand et al. | Jul 2007 | B2 |
| 7304899 | Gerstmeier et al. | Dec 2007 | B2 |
| 7308598 | Beattie et al. | Dec 2007 | B2 |
| 7310757 | Ngo et al. | Dec 2007 | B2 |
| 7328335 | Sundararajan et al. | Feb 2008 | B1 |
| 7350119 | Zuraski, Jr. et al. | Mar 2008 | B1 |
| 7383423 | Hughes et al. | Jun 2008 | B1 |
| 7415640 | Zorian et al. | Aug 2008 | B1 |
| 7529992 | Teig et al. | May 2009 | B1 |
| 7663957 | Henry et al. | Feb 2010 | B2 |
| 7702978 | Lewis et al. | Apr 2010 | B2 |
| 7710813 | Im et al. | May 2010 | B1 |
| 7757135 | Nadeau-Dostie et al. | Jul 2010 | B2 |
| 7795899 | Grohoski et al. | Sep 2010 | B1 |
| 7805766 | Christensen et al. | Sep 2010 | B2 |
| 7839707 | Askjer | Nov 2010 | B2 |
| 8112681 | Khoja et al. | Feb 2012 | B2 |
| 8194489 | Bentley et al. | Jun 2012 | B2 |
| 8242800 | Henry et al. | Aug 2012 | B2 |
| 8484543 | Anand et al. | Jul 2013 | B2 |
| 8537627 | Ouellette et al. | Sep 2013 | B2 |
| 8719648 | Gorman et al. | May 2014 | B2 |
| 8724418 | Kim et al. | May 2014 | B2 |
| 8843795 | Nakaya | Sep 2014 | B2 |
| 8879345 | Henry | Nov 2014 | B1 |
| 20040019763 | Lakhani et al. | Jan 2004 | A1 |
| 20060004942 | Hetherington et al. | Jan 2006 | A1 |
| 20060131743 | Erickson et al. | Jun 2006 | A1 |
| 20060203578 | Meaney et al. | Sep 2006 | A1 |
| 20080008015 | Darbinyan et al. | Jan 2008 | A1 |
| 20080046891 | Sanchorawala et al. | Feb 2008 | A1 |
| 20080065937 | Micheloni et al. | Mar 2008 | A1 |
| 20080184009 | Hughes et al. | Jul 2008 | A1 |
| 20080263490 | Lewis et al. | Oct 2008 | A1 |
| 20090045867 | Kwong et al. | Feb 2009 | A1 |
| 20090080232 | Chen et al. | Mar 2009 | A1 |
| 20090097335 | Taniguchi et al. | Apr 2009 | A1 |
| 20090204751 | Kushita | Aug 2009 | A1 |
| 20100229062 | Henry et al. | Sep 2010 | A1 |
| 20110035623 | Gaskins et al. | Feb 2011 | A1 |
| 20120096241 | Bell, Jr. et al. | Apr 2012 | A1 |
| 20120166763 | Henry et al. | Jun 2012 | A1 |
| 20130022951 | Hughes | Jan 2013 | A1 |
| 20130033951 | Gorman et al. | Feb 2013 | A1 |
| 20130070514 | Weiss et al. | Mar 2013 | A1 |
| 20130091312 | Ken et al. | Apr 2013 | A1 |
| 20130166944 | Park | Jun 2013 | A1 |
| 20150338904 | Henry et al. | Nov 2015 | A1 |
| 20150338905 | Henry et al. | Nov 2015 | A1 |
| 20150339231 | Henry et al. | Nov 2015 | A1 |
| Number | Date | Country |
|---|---|---|
| 1357893 | Jul 2002 | CN |
| 101180608 | May 2008 | CN |
| 101493809 | Jul 2009 | CN |
| 103377711 | Oct 2013 | CN |
| 104538059 | Apr 2015 | CN |
| 104570830 | Apr 2015 | CN |
| 104572335 | Apr 2015 | CN |
| 104575910 | Apr 2015 | CN |
| 1215682 | Jun 2002 | EP |
| Number | Date | Country | |
|---|---|---|---|
| 20150339232 A1 | Nov 2015 | US |
