Embodiments of the disclosure relate to the field of computing; and more specifically, the embodiments are related to virtual machine performance measurement.
A processor/core may enter different power states to control power consumption, where each power state corresponds to one or more specific voltages and/or frequencies. The power states include a set of performance states (commonly referred to as P-states) and idle states (commonly referred to as C-states). The number and definition of P-states and C-states vary depending on the specific processor model and architecture. Note that power states may also be referred to as power saving states/modes, energy-efficient states/modes, or other similar terms.
For example, a processor/core may be active in one of the P-states, where each P state corresponds to a level of voltage and/or frequency. P0 state is known as the highest or maximum performance state. In this state, a processor/core operates at its highest voltage and/or frequency, providing maximum processing power. P0 state is used when the system is under heavy load and requires high performance. In some embodiments, a processor/core is referred to as operating in a turbo mode herein when the processor/core enters P0 state. P1, P2, etc. are intermediate performance states that represent decreasing levels of performance. In each subsequent P-state, the processor/core operates at a lower clock frequency and voltage, reducing power consumption while sacrificing some performance. These states are used to balance performance and power consumption based on the workload. Pn is the lowest performance state, with the lowest voltage and frequency.
A set of counters may be used to measure the performance of a processor at a performance state. For example, in the x86 computer architecture, IA32_MPERF MSR and IA32_APERF MSR are a pair of mode-specific registers (MSRs) to measure processor performance at a P-state. The former Maximum Performance Counter Frequency (MPERF) MSR increments in proportion to a fixed frequency, while the latter Actual Performance Counter Frequency (APERF) MSR increments in proportion to the actual processor performance. In the ARM computer architecture, a pair of registers in the Activity Monitors Extension (AMU) are defined to measure an ARM processor performance, and they are (1) the Reference Performance Counter Register that counts cycles at a constant frequency and (2) the Delivered Performance Counter Register that counts cycles at core frequency.
Yet it is challenging to use these pairs of counters to measure performances in virtualization environment, where a client may demand measurements of a particular virtual machine (VM).
The disclosure may best be understood by referring to the following description and accompanying drawings that are used to show embodiments of the disclosure.
In the following description, numerous specific details are set forth. However, it is understood that embodiments of the disclosure 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.
Bracketed text and blocks with dashed borders (such as large dashes, small dashes, dot-dash, and dots) may be used to illustrate optional operations that add additional features to the embodiments of the disclosure. Such notation, however, should not be taken to mean that these are the only options or optional operations, and/or that blocks with solid borders are not optional in some embodiments of the disclosure.
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.
The terms “connected” means a direct electrical or magnetic connection between the things that are connected, without any intermediary devices, while the term “coupled” means either a direct electrical or magnetic connection between the things that are connected or an indirect connection through one or more passive or active intermediary devices. The term “circuit” means one or more passive and/or active components that are arranged to cooperate with one another to provide a desired function. A “set,” as used herein, refers to any positive whole number of items including one item.
As discussed, a set of counters may be used to measure the performance of a processor at a performance state. While different computer architectures have different names for the sets of counters, their functions are similar, with one counter increments in proportion to a fixed frequency, and the count so incremented is referred to as machine count (MCNT), and another counter increments in proportion to the actual processor performance and the count so incremented is referred to as actual count (ACNT). MCNT represents the number of machine cycles that have occurred on the processor since the counter was last reset, and ACNT represents the number of architectural events that have occurred on the processor since the counter was last reset. For example, the architectural events include instruction fetch, decode, and execution, memory access and write back of the processor in some embodiments.
The performance of a processor in a period may be measured using the following formula in some embodiments:
Percent_Performance=Percent_Busy*(ACNT/MCNT) (1)
where Percent_Busy represents the busy (non-idle) duration within the period and ACNT and MCNT are the respective counts within the period.
This disclosure uses Actual Performance Counter Frequency (APERF) counter and Maximum Performance Counter Frequency (MPERF) counter as exemplary counters to count ACNT and MCNT, respectively. The pair of APERF and MPERF counters may be implemented using the IA32_APERF and IA32_MPERF MSRs in the x86 computer architecture, the Reference Performance Counter Register and the Delivered Performance Counter Register in the ARM computer architecture, or another pair of counters/registers in another computer architecture, and embodiments discussed herein are not limited to a particular set of counters for performance measurements in a particular architecture.
Additionally, embodiments discussed herein are not limited to a central processing unit (CPU) based computing system, and they may be used in a computing system based on one or more graphics processing units (GPUs), accelerated processing units (APUs) each of which combines CPU and GPU capabilities on a single chip, neural-network/tensor processing units (NPUs/TPUs) each of which is an accelerator designed specifically for machine learning (ML) and/or artificial intelligence (AI) workloads, and/or other processing units. Each of these different processing units is collectively referred to as an xPU herein.
In virtualization scenarios, the need to measure performance for virtual machines (VM) has emerged. In cloud computing cases, users who rent VMs on a computer server often want to know whether the performance of the VMs they are assigned to meet expectations; and for client cases where a client computing system is used, while users of VMs enjoy the benefits of security isolation brought by virtualization, they also hope that performance testing tools based on these two APERF and MPERF counters may work in VMs.
Yet APERF and MPERF counters often do not work in virtualization scenarios. For example, IA32_APERF MSR and IA32_MPERF MSR lack virtualization support at the virtual CPU level, so that a guest operating system (also referred to as “guest” here) ran by a VM on a processor (also referred to as “host processor”) that manages and hosts the VM using a hypervisor (also referred to as a “host” to distinguish with the guest operating system) cannot use these two MSRs to measure the performance of the current virtual machine. By default, the access to the two MSRs in a VM can cause exit from the VM (VM exit). While the two MSRs can be passed through to a VM by setting an MSR bitmap to allow MSR read and write instructions (RDMSR/WRMSR) on the two MSRs without triggering any VM exit, the guest of the VM still cannot get the correct performance measurement for that VM.
The reason is that the virtual CPU (vCPU) of a VM is just the task on a hypervisor, but IA32_APERF MSR and IA32_MPERF MSR are defined as “thread” scopes, which means they are per-processor and CPU-local. IA32_APERF MSR and IA32_MPERF MSR keep increment on the hypervisor regardless of whether the vCPU thread is running. Thus, the count values read from these two MSRs include not only the absolute value of increase for the current VM, but also the count value of other workloads on the hypervisor. The performance measured for the current VM based on such counts is obviously inaccurate. Note that the hypervisor is an example of a software or firmware layer that enables the virtualization of computer hardware, which may be referred to as Virtual Machine Monitor (VMM) or another term. Embodiments of the invention are not limited to a particular way virtualization is implemented.
Two existing approaches have been implemented for performance feedback counter virtualization. One is software simulation, and the other is hardware read-only counters, each having their drawbacks as explained herein using IA32_APERF and IA32_MPERF MSRs as examples.
In this approach, the IA32_APERF and IA32_MPERF MSRs are both emulated to exclude the counting when no VMs are running. The hypervisor maintains an ACNT offset and an MCNT offset. Every time a VM exit occurs, the hypervisor will update the ACNT & MCNT offsets to make sure not to include the counting value during the VM exit handling. And every time when a VM tries to read/write IA32_APERF/IA32_MPERF MSR, a VM exit is triggered, the APERF count value is scaled with ACNT offset and the MPERF is scaled with MCNT offset and the current VM's Timestamp Counter (TSC) multiplier. The ACNT offset and MCNT offset reflect the offset of the count value of the guest's performance relative to the host, and can be used in the guest's live migration, just like the TSC offset in TSC virtualization.
Note that the TSC multiplier is also referred to as TSC scaling factor, TSC ratio, or TSC virtualization ratio, and it represents the relationship between the TSC of the virtual machine (or a VPU) and the TSC of the physical host. The TSC multiplier indicates how fast or slow the TSC of the virtual machine runs compared to the TSC of the physical processor. For example, a TSC multiple of two means that the TSC of the virtual machine increments twice as fast as the TSC of the physical host.
Software simulation of the performance feedback counters has several drawbacks. For one, the count value calculated by MSRs' save/load mechanism is inaccurate. The actual running time of the VM and the code of load/save will always have additional time loss. Additionally, software computing MPERF (scaled with TSC multiplier) always incurs overhead. And the software needs to deal with the overflow caused by the multiplication operation, which further increases the emulation overhead and increases the complexity of the software. Furthermore, if the host also needs to measure the performance of the whole computing system, the guest's write to the MSRs will break the host's count, which will seriously affect the normal operation of the host.
A computing system may support a read-only version to scale the APERF and MPERF count values in non-root mode (also referred to as guest mode), where the APERF read-only operation by a VM reads the IA32_APERF MSR of the host, the value of which may be scaled by the effective offset, and the MPERF read-only operation reads the IA32_MPERF MSR of the host, the value of which is scaled by guest's TSC multiplier. As IA32_APERF MSR and IA32_MPERF MSR are passed through to the VM, any VM read will access the read-only version. Note that for the IA32_MPERF MSR, a VM can get the correct TSC count which is scaled by VM's own TSC frequency.
Note that non-root mode is the mode of operation that is used by a VM (the corresponding guest operating system). In non-root mode, the guest operating system is restricted in its ability to access the underlying hardware. It cannot execute privileged instructions, access memory locations that are not mapped to its virtual address space, or modify certain registers. This restriction prevents guest operating systems from interfering with the hypervisor or other guest operating systems. The root mode is the privileged mode of operation that is used by the hypervisor. In the root mode, the hypervisor has complete control over the processor and all of its resources. It can execute any instruction, access any memory location, and modify any register. This level of access is necessary for the hypervisor to perform its tasks, such as scheduling virtual machines, allocating resources, and handling virtualization-specific events. The transition between root mode and non-root mode is through VM exit discussed herein above. A VM exit occurs when a guest operating system attempts to perform an operation that is not allowed in non-root mode. When a VM exit occurs, the hypervisor is notified and takes control of the processor. The hypervisor can then either emulate the instruction, trap the instruction, or allow the instruction to execute in root mode.
In the approach using the hardware read-only counters, all of the VMs and the host share the APERF and MPERF counters, and for write operations, the counter sharing will cause wrong counting. Thus, the hardware read-only counters are not a complete solution and can't support hypervisor and VM to measure performance at the same time or multiple VMs to measure performance at the same time. Additionally, when no VMs are running, the APERF and MPERF counters continue to increment the count values, and it is challenging to accurately determine VM's performance based on read-only of the APERF and MPERF counters.
Embodiments here overcome the drawbacks of the existing approaches and allow the APERF and MPERF counters to be used to accurately determine VM's performance.
Host processor 150 includes a set of performance feedback counters, including a pair of APERF counter and MPERF counter. Each counter comprises or is coupled with circuitry to monitor processing of host processor 150 and causes a count (e.g., ACNT/MCNT) to increment. The pair of APERF counter and MPERF counter to measure the performance of host processor 150 itself, regardless of the processor being in the root mode or non-root mode, may be referred to as host APERF counter and host MPERF counter respectively, and they are not shown in
Host processor 150 may include one or more logical entities (not shown), each of which represents a hardware thread or execution context within a physical CPU core. Each logical entity may be referred to as a logical processing unit (LPU), or logical xPU to indicate that the processing unit may be any of the xPUs discussed herein.
In some embodiments, each LPU comprises or is coupled to a set of performance feedback counters, including a pair of APERF and MPERF counters for that LPU. Alternatively, the LPUs may share a pair of APERF and MPERF counters. The pair of APERF counter 192 and MPERF counter 194 is shown as an example of counters for an LPU to count performance in the non-root mode (instead of APERF/MPERF counters counting regardless of the root mode or the non-root mode) The pair of APERF counter 192 and MPERF counter 194 may be referred to as a virtual APERF counter and a virtual MPERF counter, respectively to differentiate with the pair of APERF counter and MPERF counter that counts in both of the root and non-root modes. In some embodiments, the virtual APERF/MPERF counter is referred to as a vmx_aperf/vmx_mperf, where “vmx” stands for virtual machine extension. In some embodiments, APERF counter 192 and MPERF counter 194 are implemented within or coupled to a power management module/logic/circuitry 190 that manages power consumption of host processor 150.
LPUs are created by the processor itself to handle multiple threads concurrently, and they are part of the physical processor architecture and are not directly managed by the hypervisor (not shown) implemented by host processor 150. On the other hand, the hypervisor implemented by host processor 150 may manage one or more virtual processing units (VPUs) for VMs. The VPUs may also be referred to as virtual xPUs to indicate that the underlying each of the one or more processing units may be one of the xPUs discussed herein. A VM may have multiple VPUs allocated to the VM, and each VPU provides an isolated and dedicated processing unit for a VM, allowing the VM to run its own operating system and applications as if it were running on host processor 150. VPUs are part of the virtualization layer and are managed by the hypervisor, which is responsible for scheduling and allocating physical resources on host processor 150 among multiple VMs.
In some embodiments, each VPU is allocated to its own pair of APERF and MPERF counters (not shown). These per-VPU counters are referred to as the VPU virtual APERF counter and the VPU virtual MPERF counter of a VPU.
In some embodiments, a pair of APERF and MPERF counters are used to perform counter virtualization for multiple VPUs corresponding to a single LPU of host processor 150, and they are shown as APERF counter 192 and MPERF counter 194. The pair of APERF and MPERF counters 192 and 194 are used continuously by the VPUs running in turn on the LPU. The counts accumulated by the pair of APERF and MPERF counters 192 and 194 indicate the performance of the logical processing unit under non-root mode.
Both per-VPU and per-LPU counters provide useful performance measurements for a client of the virtualization environment of system. While adding a pair of virtual APERF and MPERF counters for each VPU is easier to understand, it may take too much storage space to maintain the per-VPU counters. From a high-level view, VPU is the task (software thread), and in a typical cloud scenario, there may be hundreds or thousands of VPU threads running on a logical CPU, and it may not be realistic to maintain pre-task counters for each VPU thread in this and/or other similar scenarios.
A VMCS is a data structure maintained by a hypervisor for each VPU it manages in the x86 computer architecture. The VMCS contains configuration and state information that the hypervisor uses to control the behavior of the VPU when it is running on the physical hardware. Note that while VMCS is used as an example of the data structure containing configuration and state information to control the behavior of a VPU, other data structures may be used in a different x86 (or another) computer architecture. For example, the Virtual Machine Control Register (VMCR) and Virtualization Control Register (VCR) are used in the ARM Virtualization Extension (VE) for a similar purpose in the ARM computer architecture, so are the Partition Control Block (PCB) and Partition Control Entry (PCE) in the IBM Power Architecture, where PCB/PCE contains information about the state and configuration of a virtual machine or partition. Embodiments of the invention are not limited to a particular data structure or entity used to store the configuration and state information that the hypervisor uses to control the behavior of the VPU.
In some embodiments, each VMCS (or one of VMCR/VCR or PCB/PCE in an alternative architecture) as shown contains a vmx_aperf and vmx_mperf field (shown at references 142 to 146), and details of a VMCS are explained further relating to
The execution of a VPU on an LPU of host processor 150 is initiated through a VM entry and completed through a VM exit, and VPU 104 is shown to be executed on the LPU of host processor 150 in
VM entry 132 transitions the operation from the hypervisor to VPU 104 (guest). This transition involves moving from the execution of the hypervisor to the execution of VPU 104, including loading VMCS 124, checking the reason of the prior VM exit, setting up gust state (including processor registers, flags, segment selectors, and other relevant data structures), loading guest page table that enabled the virtual-to-physical address translation for the guest, transitioning to non-root mode, executing guest instructions, handling interrupts and events, and/or monitoring and trapping instructions. During the execution of VPU 104, VMCS fields may be updated based on the execution, including one or more vmx_aperf and vmx_mperf fields 144.
VPU 104 maintains its own APERF and MPERF count values (ACNT and MCNT) through one or more vmx_aperf and vmx_mperf fields 144. Once VPU 104 starts execution, the APERF and MPERF count values are loaded into APERF counter 192 and MPERF counter 194, respectively. In some embodiments, the load may be performed through a write instruction (e.g., one that writes to an MSR, WRMSR) at reference 136. APERF counter 192 and MPERF counter 194 then count based on the loaded APERF and MPERF count values to measure the performance of VPU 104, until a VM exit event causes the execution completion of VPU 104. Upon VM exit 134, the updated count values in APERF counter 192 and MPERF counter 194 are stored to the corresponding one or more vmx_aperf and vmx_mperf fields 144. In some embodiments, the store may be performed through a read instruction (e.g., one that read to an MSR, RDMSR) at reference 138. In some embodiments and/or some scenarios (e.g., upon a VPU 104 is first initiated), instead of loading the APERF and MPERF count values (ACNT and MCNT) through one or more vmx_aperf and vmx_mperf fields 144, the load is to reset the existing count values tracked by the APERF and MPERF counters, so the APERF and MPERF counters count anew. Note that in some embodiments, the APERF and MPERF count values (ACNT and MCNT) accessible through one or more vmx_aperf and vmx_mperf fields 144 may be initialized to zeros by default in some embodiments so that the load through vmx_mperf fields 144 may achieve the same count starting points of zeros for the two counters.
The incremented counter values (from the initial loaded value to the updated count values) indicate the performance of VPU 104 within the first time period. For example, Formula (1) may be applied on the incremented count values to measure VPU 104's performance within the first time period. When VPU 104 is executed again by the LPU in a second period later, the process repeats and the APERF and MPERF count values continue incrementing, reflecting the performance of VPU 104 in both first and second periods. The counting of VPU 104 is performed only when VPU 104 is executed on the LPU, and it is not affected by any intervening execution of another VPU and may accurately reflect the performance of VPU 104 over a sustained period of time that does not require continuous execution of VPU 104 on the LPU.
Such measurement over a sustained period of time for a VPU when the VPU shares the resources (LPU) of a host processor with other VPUs. The client may understand the performance of a VM at the granularity of the VPU level (regardless of how many VMs the client deploys) and provide the client a better user experience by knowing the service level they are provided with. The measurement also provides a metric for the hypervisor to coordinate. For example, if VPU performance level at performance state P1 is determined to be low, the performance may be deemed low when based on Formula (1), the calculated performance value is below a threshold, the processor may be transitioned to operate at a lower performance state, P2, to save power for the processor.
APERF counter 192 and MPERF counter 194 are shared among multiple VPUs, including VPU 102 to VPU 106, and each VPU takes turns to use these counters (through the write and read at references 136 and 138) when the VPU is executed on the LPU. Sharing of the pair of performance feedback counters by multiple VPUs along the embodiments to scale and be used in an environment where numerous VPUs are implemented.
Note that while the discussion regarding to
Counting to measure performance of a processor uses a variety of counters of a computing system.
While host APERF counter 292 and host MPERF counter 294 may be made available for access by a VPU/VM in non-root mode 202 in some embodiments (e.g., through setting the MSR bitmap to allow MSR read and write instructions as discussed herein above), such access is unnecessary as additional virtual counters are implemented in some embodiments, shown as virtual APERF counter 282 and virtual MPERF counter 284, both of which may be implemented using respective registers (e.g., MSR registers to be accessed through RDMSR/WRMSR instruction, similar to the host APERF/MPERF counter) in some embodiments.
Virtual APERF counter 282 and virtual MPERF counter 284 may be accessed in non-root mode 202, through access instructions for APERF and MPERF count values at reference 242. The access may be through read and/or write instructions at reference 236, e.g., the RDMSR/WRMSR instruction discussed herein above. Note that without the disclosed embodiments herein, the read and/or write instructions at reference 236 would direct the access to host APERF counter 292 and host MPERF counter 294, which are known in the art. With the embodiments, the access is to be redirected to one or more vmx_aperf and vmx_mperf fields 144, through which the count values for a VPU/VM are obtained.
While the guest may read/write the count values as stored through one or more vmx_aperf and vmx_mperf fields 144, the stored count values may be scaled through a scaling module 246 when they are read from or written to (at reference 262) virtual APERF counter 282 and virtual MPERF counter 284.
For example, at a VM entry, the count values as stored through one or more vmx_aperf and vmx_mperf fields 144 are written to virtual APERF counter 282 and virtual MPERF counter 284, the write/load does not need scaling by the scaling module 246 as the count values are the starting values to count. At a VM exit, however, the updated count values may need to be scaled by scaling module 246 so that the read/stored updated count values are align with the TSC multiplier or count offset (ACNT offset or MCNT offset) of the corresponding VPU. The scaling may be performed by multiplying count value with the TSC multiplier which reflects VM's TSC virtualization, or offsetting the count offset to reflect the performance offset of the VPU relative to the PU.
Note that ACNT offset and MCNT offset are optional, and their existence depends on the specific implementation of the underlying virtual APERF and MPERF counters. If virtual APERF and MPERF counters are implemented based on offsetting, that is, they implement the continuously growing counters, and the starting value of a count is marked by offsetting, then reading ACNT offset and MCNT offset through vmx_aperf and vmx_mperf fields 144 is needed. These ACNF/MCNT offsets may be used to adjust the count values in a guest live migration case in some embodiments. For example, on a current VPU, the VPU's ACNT count (a0) and MCNT count (x0) start from the ACNT offset (b0) and MCNT offset (y0) relative to the VPU's virtual APERF counter (count value: c0) and MPERF counter (count value: z0), and the corresponding formulas are: a0+b0=c0 and x0+y0=z0. When the VPU migrates to another VPU on different hardware machine (e.g., a difference core), the ACNT offset and MCNT offset need to be adjusted to align with the new count values of the virtual APERF counter (c1) and MPERF counter (z1) of the new VPU to ensure continuity of the APERF and MPERF counts of the VPU, then the new ACNT offsets is b1=c1−c0+b0, and the new MCNT offset is y1=z1−z0+y0, so that the Guest's APERF count on new machine equals the value on previous machine: a1=c1−b1=c0−b0=a0, and the Guest's MPERF count on new machine also equals the value on previous machine: x1=z1−y1=z0−y0=x0. With offsets adjustment on virtual APERF and MPERF counters, this approach guarantees continuous performance counting when VPU migrates between different hardware machines. If virtual APERF and MPERF counters aren't based on offsetting mechanisms, then ACNT offset and MCNT offset are not needed.
Through using the scaling module 246 to scale the values counted in virtual APERF counter 282 and virtual MPERF counter 284, VPUs and VMs with different TSC multipliers may share the same pair of virtual APERF counter 282 and virtual MPERF counter 284. The sharing of virtual APERF counter 282 and virtual MPERF counter 284 among all the VPUs/VMs with different TSC multipliers allow the virtual machine performance measurement disclosed herein scalable to any number of VPUs/VMs to be supported by a computing system.
A VPU (e.g., VPU 104) maintains its own APERF and MPERF count values (ACNT and MCNT) through its own set of vmx_aperf and vmx_mperf fields (e.g., one or more vmx_aperf and vmx_mperf fields 144), which may be implemented in a VMCS (e.g., VMCS 124) or another data structure as discussed herein above.
VMCS 124 may include a guest-state area 302 that contains fields related to the state of the virtual machine as seen by the guest operating system. Examples include registers, segment selectors, and flags. VMCS 124 may include a host-state area 304 that stores the values of processor state components for the host system (e.g., host processor 150). These components include host general-purpose registers, segment selectors, and control registers. VMCS 124 may further include VM-execution control fields 306 that contain fields that control the behavior of the VM/VPU. These fields include settings for interrupt handling, virtual-machine exit reasons, and various other VM-execution control fields. VMCS 124 may also include VM-exit control fields 308 that control VM/VPU exits, VM-entry control fields 310 that control VM/VPU entries, and VM-exit information field 312 receive information on VM exits and describe the cause and the nature of VM exits (read only in some embodiments).
In some embodiments, the one or more vmx_aperf and vmx_mperf fields 144 are fields within VM-execution control fields 306. The one or more vmx_aperf and vmx_mperf fields 144 may include an enablement indication regarding whether the virtual APERF and MPERF counters for VPU/VM is enabled for the virtualization of a processor, shown as the vmx_aperf and vmx_mperf virtualization enable field 322. In some embodiments, field 322 is a set of control bits, e.g., a single bit where a value of one indicates enablement and a value of zero indicates disablement, or vice versa.
The one or more vmx_aperf and vmx_mperf fields 144 may further include a virtual APERF and MPERF address field 324 to indicate where to find the APERF and MPERF count values. In some embodiments, the virtual APERF and MPERF address field 324 indicates the (physical/virtual) address of a memory page, which points to the location that stores the virtual APERF count value 326 (vmx_aperf value) and virtual MPERF count value 328 (vmx_mperf value). In some embodiments, the virtual APERF count offset 336 (vmx_aperf offset) and the virtual MPERF count offset 338 (vmx_mperf offset) are in the location for scaling as well. Once an instruction is received to read from or write to an APERF/MPERF count value of a VPU/VM, the read or write will be redirected to the corresponding virtual APERF count value 326 (vmx_aperf value) and virtual MPERF count value 328 of the VPU/VM.
At reference 402, responsive to a virtual machine entry to a virtual processing unit of a processor, the circuitry writes a first count value and a second count value corresponding to the virtual processing unit to a first counter and a second counter of the processor, respectively. The first and second counters are the virtual MPERF and virtual APERF counters, respectively.
At reference 404, the first counter is incremented based on the first count value at a fixed frequency, and the second counter is incremented based on the second count value in proportion of architectural event occurrence of the virtual processing unit. The architectural event occurrence of the virtual processing unit indicates the number of architectural events that have occurred since the virtual machine entry to the virtual processing unit of the processor as the first counter increment at the fixed frequency.
At reference 406, responsive to a virtual machine exit from the virtual processing unit, the circuitry reads a resulting first count value and a resulting second count value from the first and second counters respectively to determine a measure of performance of the virtual processing unit within a corresponding duration between the virtual machine entry and the virtual machine exit.
In some embodiments, the resulting first and second count values indicate the performance of the virtual processing unit at a corresponding performance state of the processor that maps to one or more frequency and voltage settings.
In some embodiments, the first and second count values are obtained based on a first value and a second value stored through a virtual machine control structure corresponding to the virtual processing unit. In some embodiments, the first and second values are stored in through a virtual machine execution control field of the virtual machine control structure.
In some embodiments, one or more of the resulting first and second count values are scaled based on a time-stamp counter (TSC) multiplier or count offset of the virtual processing unit prior to storing as updates of the first and second values through the virtual machine execution control field. The scaling at the VM exit is discussed in further details relating to
In some embodiments, an access request to the first counter or a second counter of the processor is redirected to a corresponding value of the first and second values stored in the virtual machine control structure.
In some embodiments, the measure of performance of the virtual processing unit is determined based on a ratio between the resulting first and second count values from the first and second counter of the processor. For example, Formula (1) may be used to determine the performance of the virtual processing unit as discussed herein above.
In some embodiments, the first and second count values corresponding to the virtual processing unit are written to a first register and a second register corresponding to the first and second counters, respectively, and the resulting first and second count values are read from the first and second registers, respectively. The first and second registers are for the virtual MPERF counter and virtual APERF counters, respectively.
In some embodiments, a set of bits corresponding to the first and second registers is set to enable access to the first and second registers prior to accessing the first and second registers in a non-root mode. In some embodiments, the set of bits is in the vmx_aperf and vmx_mperf virtualization enable field 322.
In some embodiments, a set of bitmap bits corresponding to the first and second registers is set to enable access to the first and second registers prior to accessing the first and second registers in a non-root mode.
In some embodiments, the first and second registers are separated from registers to measure the performance of the processor in both root mode and non-root mode. The registers to measure the performance of the processor are host APERF and MPERF MSR registers in some embodiments.
Embodiments disclosed herein may be applied to computing systems in a virtualized environment, and the computing systems may be in a variety of computer architectures and implement a variety of xPUs, and
Processors 570 and 580 are shown including integrated memory controller (IMC) circuitry 572 and 582, respectively. Processor 570 also includes interface circuits 576 and 578; similarly, second processor 580 includes interface circuits 586 and 588. Processors 570, 580 may exchange information via the interface 550 using interface circuits 578, 588. IMCs 572 and 582 couple the processors 570, 580 to respective memories, namely a memory 532 and a memory 534, which may be portions of main memory locally attached to the respective processors.
Processors 570, 580 may each exchange information with a network interface (NW I/F) 590 via individual interfaces 552, 554 using interface circuits 576, 594, 586, 598. The network interface 590 (e.g., one or more of an interconnect, bus, and/or fabric, and in some examples is a chipset) may optionally exchange information with a coprocessor 538 via an interface circuit 592. In some examples, the coprocessor 538 is a special-purpose processor, such as, for example, a high-throughput processor, a network or communication processor, compression engine, graphics processor, general purpose graphics processing unit (GPGPU), neural-network processing unit (NPU), embedded processor, or the like.
A shared cache (not shown) may be included in either processor 570, 580 or outside of both processors, yet connected with the processors via an interface such as 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.
Network interface 590 may be coupled to a first interface 516 via interface circuit 596. In some examples, first interface 516 may be an interface such as a Peripheral Component Interconnect (PCI) interconnect, a PCI Express interconnect or another I/O interconnect. In some examples, first interface 516 is coupled to a power control unit (PCU) 517, which may include circuitry, software, and/or firmware to perform power management operations with regard to the processors 570, 580 and/or coprocessor 538. PCU 517 provides control information to a voltage regulator (not shown) to cause the voltage regulator to generate the appropriate regulated voltage. PCU 517 also provides control information to control the operating voltage generated. In various examples, PCU 517 may include a variety of power management logic units (circuitry) to perform hardware-based power management. Such power management may be wholly processor controlled (e.g., by various processor hardware, and which may be triggered by workload and/or power, thermal or other processor constraints) and/or the power management may be performed responsive to external sources (such as a platform or power management source or system software).
PCU 517 is illustrated as being present as logic separate from the processor 570 and/or processor 580. In other cases, PCU 517 may execute on a given one or more of cores (not shown) of processor 570 or 580. In some cases, PCU 517 may be implemented as a microcontroller (dedicated or general-purpose) or other control logic configured to execute its own dedicated power management code, sometimes referred to as P-code. In yet other examples, power management operations to be performed by PCU 517 may be implemented externally to a processor, such as by way of a separate power management integrated circuit (PMIC) or another component external to the processor. In yet other examples, power management operations to be performed by PCU 517 may be implemented within BIOS or other system software.
Various I/O devices 514 may be coupled to first interface 516, along with a bus bridge 518 which couples first interface 516 to a second interface 520. In some examples, one or more additional processor(s) 515, such as coprocessors, high throughput many integrated core (MIC) processors, GPGPUs, accelerators (such as graphics accelerators or digital signal processing (DSP) units), field programmable gate arrays (FPGAs), or any other processor, are coupled to first interface 516. In some examples, second interface 520 may be a low pin count (LPC) interface. Various devices may be coupled to second interface 520 including, for example, a keyboard and/or mouse 522, communication devices 527 and storage circuitry 528. Storage circuitry 528 may be one or more non-transitory machine-readable storage media as described below, such as a disk drive or other mass storage device which may include instructions/code and data 530 and may implement the storage 'ISAB03 in some examples. Further, an audio I/O 524 may be coupled to second interface 520. Note that other architectures than the point-to-point architecture described above are possible. For example, instead of the point-to-point architecture, a system such as multiprocessor system 500 may implement a multi-drop interface or other such architecture.
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) computing. Such different processors lead to different computing 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 (SoC) that may be included on the same die as the described CPU (sometimes referred to as the application core(s) or application processor(s)), the above described coprocessor, and additional functionality. Example core architectures are described next, followed by descriptions of example processors and computer architectures.
Thus, different implementations of the processor 600 may include: 1) a CPU with the special purpose logic 608 being integrated graphics and/or scientific (throughput) logic (which may include one or more cores, not shown), and the cores 602(A)-(N) being one or more general purpose cores (e.g., general purpose in-order cores, general purpose out-of-order cores, or a combination of the two); 2) a coprocessor with the cores 602(A)-(N) being a large number of special purpose cores intended primarily for graphics and/or scientific (throughput); and 3) a coprocessor with the cores 602(A)-(N) being a large number of general purpose in-order cores. Thus, the processor 600 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 600 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, complementary metal oxide semiconductor (CMOS), bipolar CMOS (BiCMOS), P-type metal oxide semiconductor (PMOS), or N-type metal oxide semiconductor (NMOS).
A memory hierarchy includes one or more levels of cache unit(s) circuitry 604(A)-(N) within the cores 602(A)-(N), a set of one or more shared cache unit(s) circuitry 606, and external memory (not shown) coupled to the set of integrated memory controller unit(s) circuitry 614. The set of one or more shared cache unit(s) circuitry 606 may include one or more mid-level caches, such as level 2 (L2), level 3 (L3), level 4 (L4), or other levels of cache, such as a last level cache (LLC), and/or combinations thereof. While in some examples interface network circuitry 612 (e.g., a ring interconnect) interfaces the special purpose logic 608 (e.g., integrated graphics logic), the set of shared cache unit(s) circuitry 606, and the system agent unit circuitry 610, alternative examples use any number of well-known techniques for interfacing such units. In some examples, coherency is maintained between one or more of the shared cache unit(s) circuitry 606 and cores 602(A)-(N). In some examples, interface controller units circuitry 616 couple the cores 602 to one or more other devices 618 such as one or more I/O devices, storage, one or more communication devices (e.g., wireless networking, wired networking, etc.), etc.
In some examples, one or more of the cores 602(A)-(N) are capable of multi-threading. The system agent unit circuitry 610 includes those components coordinating and operating cores 602(A)-(N). The system agent unit circuitry 610 may include, for example, power control unit (PCU) circuitry and/or display unit circuitry (not shown). The PCU may be or may include logic and components needed for regulating the power state of the cores 602(A)-(N) and/or the special purpose logic 608 (e.g., integrated graphics logic). The display unit circuitry is for driving one or more externally connected displays.
The cores 602(A)-(N) may be homogenous in terms of instruction set architecture (ISA). Alternatively, the cores 602(A)-(N) may be heterogeneous in terms of ISA; that is, a subset of the cores 602(A)-(N) may be capable of executing an ISA, while other cores may be capable of executing only a subset of that ISA or another ISA.
The processing subsystem 701, for example, includes one or more parallel processor(s) 712 coupled to memory hub 705 via a bus or other communication link 713. The communication link 713 may be one of any number of standards-based communication link technologies or protocols, such as, but not limited to PCI Express, or may be a vendor specific communications interface or communications fabric. The one or more parallel processor(s) 712 may form a computationally focused parallel or vector processing system that can include a large number of processing cores and/or processing clusters, such as a many integrated core (MIC) processor. For example, the one or more parallel processor(s) 712 form a graphics processing subsystem that can output pixels to one of the one or more display device(s) 710A coupled via the I/O hub 707. The one or more parallel processor(s) 712 can also include a display controller and display interface (not shown) to enable a direct connection to one or more display device(s) 710B.
Within the I/O subsystem 711, a system storage unit 714 can connect to the I/O hub 707 to provide a storage mechanism for the computing system 700. An I/O switch 716 can be used to provide an interface mechanism to enable connections between the I/O hub 707 and other components, such as a network adapter 718 and/or wireless network adapter 719 that may be integrated into the platform, and various other devices that can be added via one or more add-in device(s) 720. The add-in device(s) 720 may also include, for example, one or more external graphics processor devices, graphics cards, and/or compute accelerators. The network adapter 718 can be an Ethernet adapter or another wired network adapter. The wireless network adapter 719 can include one or more of a Wi-Fi, Bluetooth, near field communication (NFC), or other network device that includes one or more wireless radios.
The computing system 700 can include other components not explicitly shown, including USB or other port connections, optical storage drives, video capture devices, and the like, which may also be connected to the I/O hub 707. Communication paths interconnecting the various components in
The one or more parallel processor(s) 712 may incorporate circuitry optimized for graphics and video processing, including, for example, video output circuitry, and constitutes a graphics processing unit (GPU). Alternatively or additionally, the one or more parallel processor(s) 712 can incorporate circuitry optimized for general purpose processing, while preserving the underlying computational architecture, described in greater detail herein. Components of the computing system 700 may be integrated with one or more other system elements on a single integrated circuit. For example, the one or more parallel processor(s) 712, memory hub 705, processor(s) 702, and I/O hub 707 can be integrated into a system on chip (SoC) integrated circuit. Alternatively, the components of the computing system 700 can be integrated into a single package to form a system in package (SIP) configuration. In some examples at least a portion of the components of the computing system 700 can be integrated into a multi-chip module (MCM), which can be interconnected with other multi-chip modules into a modular computing system.
It will be appreciated that the computing system 700 shown herein is illustrative and that variations and modifications are possible. The connection topology, including the number and arrangement of bridges, the number of processor(s) 702, and the number of parallel processor(s) 712, may be modified as desired. For instance, system memory 704 can be connected to the processor(s) 702 directly rather than through a bridge, while other devices communicate with system memory 704 via the memory hub 705 and the processor(s) 702. In other alternative topologies, the parallel processor(s) 712 are connected to the I/O hub 707 or directly to one of the one or more processor(s) 702, rather than to the memory hub 705. In other examples, the I/O hub 707 and memory hub 705 may be integrated into a single chip. It is also possible that two or more sets of processor(s) 702 are attached via multiple sockets, which can couple with two or more instances of the parallel processor(s) 712.
Some of the particular components shown herein are optional and may not be included in all implementations of the computing system 700. For example, any number of add-in cards or peripherals may be supported, or some components may be eliminated. Furthermore, some architectures may use different terminology for components similar to those illustrated in
The parallel processor 800 includes a parallel processing unit 802. The parallel processing unit includes an I/O unit 804 that enables communication with other devices, including other instances of the parallel processing unit 802. The I/O unit 804 may be directly connected to other devices. For instance, the I/O unit 804 connects with other devices via the use of a hub or switch interface, such as memory hub 705. The connections between the memory hub 705 and the I/O unit 804 form a communication link 713. Within the parallel processing unit 802, the I/O unit 804 connects with a host interface 806 and a memory crossbar 816, where the host interface 806 receives commands directed to performing processing operations and the memory crossbar 816 receives commands directed to performing memory operations.
When the host interface 806 receives a command buffer via the I/O unit 804, the host interface 806 can direct work operations to perform those commands to a front end 808. In some examples the front end 808 couples with a scheduler 810, which is configured to distribute commands or other work items to a processing cluster array 812. The scheduler 810 ensures that the processing cluster array 812 is properly configured and in a valid state before tasks are distributed to the processing clusters of the processing cluster array 812. The scheduler 810 may be implemented via firmware logic executing on a microcontroller. The microcontroller implemented scheduler 810 is configurable to perform complex scheduling and work distribution operations at coarse and fine granularity, enabling rapid preemption and context switching of threads executing on the processing cluster array 812. Preferably, the host software can prove workloads for scheduling on the processing cluster array 812 via one of multiple graphics processing doorbells. In other examples, polling for new workloads or interrupts can be used to identify or indicate availability of work to perform. The workloads can then be automatically distributed across the processing cluster array 812 by the scheduler 810 logic within the scheduler microcontroller.
The processing cluster array 812 can include up to “N” processing clusters (e.g., cluster 814A, cluster 814B, through cluster 814N). Each cluster 814A-814N of the processing cluster array 812 can execute a large number of concurrent threads. The scheduler 810 can allocate work to the clusters 814A-814N of the processing cluster array 812 using various scheduling and/or work distribution algorithms, which may vary depending on the workload arising for each type of program or computation. The scheduling can be handled dynamically by the scheduler 810 or can be assisted in part by compiler logic during compilation of program logic configured for execution by the processing cluster array 812. Optionally, different clusters 814A-814N of the processing cluster array 812 can be allocated for processing different types of programs or for performing different types of computations.
The processing cluster array 812 can be configured to perform various types of parallel processing operations. For example, the processing cluster array 812 is configured to perform general-purpose parallel compute operations. For example, the processing cluster array 812 can include logic to execute processing tasks including filtering of video and/or audio data, performing modeling operations, including physics operations, and performing data transformations.
The processing cluster array 812 is configured to perform parallel graphics processing operations. In such examples in which the parallel processor 800 is configured to perform graphics processing operations, the processing cluster array 812 can include additional logic to support the execution of such graphics processing operations, including, but not limited to texture sampling logic to perform texture operations, as well as tessellation logic and other vertex processing logic. Additionally, the processing cluster array 812 can be configured to execute graphics processing related shader programs such as, but not limited to vertex shaders, tessellation shaders, geometry shaders, and pixel shaders. The parallel processing unit 802 can transfer data from system memory via the I/O unit 804 for processing. During processing the transferred data can be stored to on-chip memory (e.g., parallel processor memory 822) during processing, then written back to system memory.
In examples in which the parallel processing unit 802 is used to perform graphics processing, the scheduler 810 may be configured to divide the processing workload into approximately equal sized tasks, to better enable distribution of the graphics processing operations to multiple clusters 814A-814N of the processing cluster array 812. In some of these examples, portions of the processing cluster array 812 can be configured to perform different types of processing. For example, a first portion may be configured to perform vertex shading and topology generation, a second portion may be configured to perform tessellation and geometry shading, and a third portion may be configured to perform pixel shading or other screen space operations, to produce a rendered image for display. Intermediate data produced by one or more of the clusters 814A-814N may be stored in buffers to allow the intermediate data to be transmitted between clusters 814A-814N for further processing.
During operation, the processing cluster array 812 can receive processing tasks to be executed via the scheduler 810, which receives commands defining processing tasks from front end 808. For graphics processing operations, processing tasks can include indices of data to be processed, e.g., surface (patch) data, primitive data, vertex data, and/or pixel data, as well as state parameters and commands defining how the data is to be processed (e.g., what program is to be executed). The scheduler 810 may be configured to fetch the indices corresponding to the tasks or may receive the indices from the front end 808. The front end 808 can be configured to ensure the processing cluster array 812 is configured to a valid state before the workload specified by incoming command buffers (e.g., batch-buffers, push buffers, etc.) is initiated.
Each of the one or more instances of the parallel processing unit 802 can couple with parallel processor memory 822. The parallel processor memory 822 can be accessed via the memory crossbar 816, which can receive memory requests from the processing cluster array 812 as well as the I/O unit 804. The memory crossbar 816 can access the parallel processor memory 822 via a memory interface 818. The memory interface 818 can include multiple partition units (e.g., partition unit 820A, partition unit 820B, through partition unit 820N) that can each couple to a portion (e.g., memory unit) of parallel processor memory 822. The number of partition units 820A-820N may be configured to be equal to the number of memory units, such that a first partition unit 820A has a corresponding first memory unit 824A, a second partition unit 820B has a corresponding second memory unit 824B, and an Nth partition unit 820N has a corresponding Nth memory unit 824N. In other examples, the number of partition units 820A-820N may not be equal to the number of memory devices.
The memory units 824A-824N can include various types of memory devices, including dynamic random-access memory (DRAM) or graphics random access memory, such as synchronous graphics random access memory (SGRAM), including graphics double data rate (GDDR) memory. Optionally, the memory units 824A-824N may also include 3D stacked memory, including but not limited to high bandwidth memory (HBM). Persons skilled in the art will appreciate that the specific implementation of the memory units 824A-824N can vary and can be selected from one of various conventional designs. Render targets, such as frame buffers or texture maps may be stored across the memory units 824A-824N, allowing partition units 820A-820N to write portions of each render target in parallel to efficiently use the available bandwidth of parallel processor memory 822. In some examples, a local instance of the parallel processor memory 822 may be excluded in favor of a unified memory design that utilizes system memory in conjunction with local cache memory.
Optionally, any one of the clusters 814A-814N of the processing cluster array 812 has the ability to process data that will be written to any of the memory units 824A-824N within parallel processor memory 822. The memory crossbar 816 can be configured to transfer the output of each cluster 814A-814N to any partition unit 820A-820N or to another cluster 814A-814N, which can perform additional processing operations on the output. Each cluster 814A-814N can communicate with the memory interface 818 through the memory crossbar 816 to read from or write to various external memory devices. In one of the examples with the memory crossbar 816 the memory crossbar 816 has a connection to the memory interface 818 to communicate with the I/O unit 804, as well as a connection to a local instance of the parallel processor memory 822, enabling the processing units within the different processing clusters 814A-814N to communicate with system memory or other memory that is not local to the parallel processing unit 802. Generally, the memory crossbar 816 may, for example, be able to use virtual channels to separate traffic streams between the clusters 814A-814N and the partition units 820A-820N.
While a single instance of the parallel processing unit 802 is illustrated within the parallel processor 800, any number of instances of the parallel processing unit 802 can be included. For example, multiple instances of the parallel processing unit 802 can be provided on a single add-in card, or multiple add-in cards can be interconnected. For example, the parallel processor 800 can be an add-in device, such as add-in device 720 of
In some examples, the parallel processing unit 802 can be partitioned into multiple instances. Those multiple instances can be configured to execute workloads associated with different clients in an isolated manner, enabling a pre-determined quality of service to be provided for each client. For example, each cluster 814A-814N can be compartmentalized and isolated from other clusters, allowing the processing cluster array 812 to be divided into multiple compute partitions or instances. In such configuration, workloads that execute on an isolated partition are protected from faults or errors associated with a different workload that executes on a different partition. The partition units 820A-820N can be configured to enable a dedicated and/or isolated path to memory for the clusters 814A-814N associated with the respective compute partitions. This datapath isolation enables the compute resources within a partition can communicate with one or more assigned memory units 824A-824N without being subjected to inference by the activities of other partitions.
In graphics applications, the ROP 826 is a processing unit that performs raster operations such as stencil, z test, blending, and the like. The ROP 826 then outputs processed graphics data that is stored in graphics memory. In some examples the ROP 826 includes or couples with a CODEC 827 that includes compression logic to compress depth or color data that is written to memory or the L2 cache 821 and decompress depth or color data that is read from memory or the L2 cache 821. The compression logic can be lossless compression logic that makes use of one or more of multiple compression algorithms. The type of compression that is performed by the CODEC 827 can vary based on the statistical characteristics of the data to be compressed. For example, in some examples, delta color compression is performed on depth and color data on a per-tile basis. In some examples the CODEC 827 includes compression and decompression logic that can compress and decompress compute data associated with machine learning operations. The CODEC 827 can, for example, compress sparse matrix data for sparse machine learning operations. The CODEC 827 can also compress sparse matrix data that is encoded in a sparse matrix format (e.g., coordinate list encoding (COO), compressed sparse row (CSR), compress sparse column (CSC), etc.) to generate compressed and encoded sparse matrix data. The compressed and encoded sparse matrix data can be decompressed and/or decoded before being processed by processing elements or the processing elements can be configured to consume compressed, encoded, or compressed and encoded data for processing.
The ROP 826 may be included within each processing cluster (e.g., cluster 814A-814N of
Operation of the processing cluster 814 can be controlled via a pipeline manager 832 that distributes processing tasks to SIMT parallel processors. The pipeline manager 832 receives instructions from the scheduler 810 of
Each graphics multiprocessor 834 within the processing cluster 814 can include an identical set of functional execution logic (e.g., arithmetic logic units, load-store units, etc.). The functional execution logic can be configured in a pipelined manner in which new instructions can be issued before previous instructions are complete. The functional execution logic supports a variety of operations including integer and floating-point arithmetic, comparison operations, Boolean operations, bit-shifting, and computation of various algebraic functions. The same functional-unit hardware could be leveraged to perform different operations and any combination of functional units may be present.
The instructions transmitted to the processing cluster 814 constitute a thread. A set of threads executing across the set of parallel processing engines is a thread group. A thread group executes the same program on different input data. Each thread within a thread group can be assigned to a different processing engine within a graphics multiprocessor 834. A thread group may include fewer threads than the number of processing engines within the graphics multiprocessor 834. When a thread group includes fewer threads than the number of processing engines, one or more of the processing engines may be idle during cycles in which that thread group is being processed. A thread group may also include more threads than the number of processing engines within the graphics multiprocessor 834. When the thread group includes more threads than the number of processing engines within the graphics multiprocessor 834, processing can be performed over consecutive clock cycles. Optionally, multiple thread groups can be executed concurrently on the graphics multiprocessor 834.
The graphics multiprocessor 834 may include an internal cache memory to perform load and store operations. Optionally, the graphics multiprocessor 834 can forego an internal cache and use a cache memory (e.g., level 1 (L1) cache 848) within the processing cluster 814. Each graphics multiprocessor 834 also has access to level 2 (L2) caches within the partition units (e.g., partition units 820A-820N of
Each processing cluster 814 may include an MMU 845 (memory management unit) that is configured to map virtual addresses into physical addresses. In other examples, one or more instances of the MMU 845 may reside within the memory interface 818 of
In graphics and computing applications, a processing cluster 814 may be configured such that each graphics multiprocessor 834 is coupled to a texture unit 836 for performing texture mapping operations, e.g., determining texture sample positions, reading texture data, and filtering the texture data. Texture data is read from an internal texture L1 cache (not shown) or in some examples from the L1 cache within graphics multiprocessor 834 and is fetched from an L2 cache, local parallel processor memory, or system memory, as needed. Each graphics multiprocessor 834 outputs processed tasks to the data crossbar 840 to provide the processed task to another processing cluster 814 for further processing or to store the processed task in an L2 cache, local parallel processor memory, or system memory via the memory crossbar 816. A preROP 842 (pre-raster operations unit) is configured to receive data from graphics multiprocessor 834, direct data to ROP units, which may be located with partition units as described herein (e.g., partition units 820A-820N of
It will be appreciated that the core architecture described herein is illustrative and that variations and modifications are possible. Any number of processing units, e.g., graphics multiprocessor 834, texture units 836, preROPs 842, etc., may be included within a processing cluster 814. Further, while only one processing cluster 814 is shown, a parallel processing unit as described herein may include any number of instances of the processing cluster 814. Optionally, each processing cluster 814 can be configured to operate independently of other processing clusters 814 using separate and distinct processing units, L1 caches, L2 caches, etc.
The instruction cache 852 may receive a stream of instructions to execute from the pipeline manager 832. The instructions are cached in the instruction cache 852 and dispatched for execution by the instruction unit 854. The instruction unit 854 can dispatch instructions as thread groups (e.g., warps), with each thread of the thread group assigned to a different execution unit within GPGPU core 862. An instruction can access any of a local, shared, or global address space by specifying an address within a unified address space. The address mapping unit 856 can be used to translate addresses in the unified address space into a distinct memory address that can be accessed by the load/store units 866.
The register file 858 provides a set of registers for the functional units of the graphics multiprocessor 834. The register file 858 provides temporary storage for operands connected to the data paths of the functional units (e.g., GPGPU cores 862, load/store units 866) of the graphics multiprocessor 834. The register file 858 may be divided between each of the functional units such that each functional unit is allocated a dedicated portion of the register file 858. For example, the register file 858 may be divided between the different warps being executed by the graphics multiprocessor 834.
The GPGPU cores 862 can each include floating point units (FPUs) and/or integer arithmetic logic units (ALUs) that are used to execute instructions of the graphics multiprocessor 834. In some implementations, the GPGPU cores 862 can include hardware logic that may otherwise reside within the tensor and/or ray-tracing cores 863. The GPGPU cores 862 can be similar in architecture or can differ in architecture. For example and in some examples, a first portion of the GPGPU cores 862 include a single precision FPU and an integer ALU while a second portion of the GPGPU cores include a double precision FPU. Optionally, the FPUs can implement the IEEE 754-2008 standard for floating point arithmetic or enable variable precision floating point arithmetic. The graphics multiprocessor 834 can additionally include one or more fixed function or special function units to perform specific functions such as copy rectangle or pixel blending operations. One or more of the GPGPU cores can also include fixed or special function logic.
The GPGPU cores 862 may include SIMD logic capable of performing a single instruction on multiple sets of data. Optionally, GPGPU cores 862 can physically execute SIMD4, SIMD8, and SIMD16 instructions and logically execute SIMD1, SIMD2, and SIMD32 instructions. The SIMD instructions for the GPGPU cores can be generated at compile time by a shader compiler or automatically generated when executing programs written and compiled for single program multiple data (SPMD) or SIMT architectures. Multiple threads of a program configured for the SIMT execution model can be executed via a single SIMD instruction. For example and in some examples, eight SIMT threads that perform the same or similar operations can be executed in parallel via a single SIMD8 logic unit.
The memory and cache interconnect 868 is an interconnect network that connects each of the functional units of the graphics multiprocessor 834 to the register file 858 and to the shared memory 870. For example, the memory and cache interconnect 868 is a crossbar interconnect that allows the load/store unit 866 to implement load and store operations between the shared memory 870 and the register file 858. The register file 858 can operate at the same frequency as the GPGPU cores 862, thus data transfer between the GPGPU cores 862 and the register file 858 is very low latency. The shared memory 870 can be used to enable communication between threads that execute on the functional units within the graphics multiprocessor 834. The cache memory 872 can be used as a data cache for example, to cache texture data communicated between the functional units and the texture unit 836. The shared memory 870 can also be used as a program managed cached. The shared memory 870 and the cache memory 872 can couple with the data crossbar 840 to enable communication with other components of the processing cluster. Threads executing on the GPGPU cores 862 can programmatically store data within the shared memory in addition to the automatically cached data that is stored within the cache memory 872.
The graphics multiprocessor 925 of
The various components can communicate via an interconnect fabric 927. The interconnect fabric 927 may include one or more crossbar switches to enable communication between the various components of the graphics multiprocessor 925. The interconnect fabric 927 may be a separate, high-speed network fabric layer upon which each component of the graphics multiprocessor 925 is stacked. The components of the graphics multiprocessor 925 communicate with remote components via the interconnect fabric 927. For example, the cores 936A-936B, 937A-937B, and 938A-938B can each communicate with shared memory 946 via the interconnect fabric 927. The interconnect fabric 927 can arbitrate communication within the graphics multiprocessor 925 to ensure a fair bandwidth allocation between components.
The graphics multiprocessor 950 of
Persons skilled in the art will understand that the architecture described in
The parallel processor or GPGPU as described herein may be communicatively coupled to host/processor cores to accelerate graphics operations, machine-learning operations, pattern analysis operations, and various general-purpose GPU (GPGPU) functions. The GPU may be communicatively coupled to the host processor/cores over a bus or other interconnect (e.g., a high-speed interconnect such as PCIe, NVLink, or other known protocols, standardized protocols, or proprietary protocols). In other examples, the GPU may be integrated on the same package or chip as the cores and communicatively coupled to the cores over an internal processor bus/interconnect (i.e., internal to the package or chip). Regardless of the manner in which the GPU is connected, the processor cores may allocate work to the GPU in the form of sequences of commands/instructions contained in a work descriptor. The GPU then uses dedicated circuitry/logic for efficiently processing these commands/instructions.
As illustrated, a multi-core group 965A may include a set of graphics cores 970, a set of tensor cores 971, and a set of ray tracing cores 972. A scheduler/dispatcher 968 schedules and dispatches the graphics threads for execution on the various cores 970, 971, 972. A set of register files 969 store operand values used by the cores 970, 971, 972 when executing the graphics threads. These may include, for example, integer registers for storing integer values, floating point registers for storing floating point values, vector registers for storing packed data elements (integer and/or floating-point data elements) and tile registers for storing tensor/matrix values. The tile registers may be implemented as combined sets of vector registers.
One or more combined level 1 (L1) caches and shared memory units 973 store graphics data such as texture data, vertex data, pixel data, ray data, bounding volume data, etc., locally within each multi-core group 965A. One or more texture units 974 can also be used to perform texturing operations, such as texture mapping and sampling. A Level 2 (L2) cache 975 shared by all or a subset of the multi-core groups 965A-965N stores graphics data and/or instructions for multiple concurrent graphics threads. As illustrated, the L2 cache 975 may be shared across a plurality of multi-core groups 965A-965N. One or more memory controllers 967 couple the GPU 980 to a memory 966 which may be a system memory (e.g., DRAM) and/or a dedicated graphics memory (e.g., GDDR6 memory).
Input/output (I/O) circuitry 963 couples the GPU 980 to one or more I/O devices 962 such as digital signal processors (DSPs), network controllers, or user input devices. An on-chip interconnect may be used to couple the I/O devices 962 to the GPU 980 and memory 966. One or more I/O memory management units (IOMMUs) 964 of the I/O circuitry 963 couple the I/O devices 962 directly to the system memory 966. Optionally, the IOMMU 964 manages multiple sets of page tables to map virtual addresses to physical addresses in system memory 966. The I/O devices 962, CPU(s) 961, and GPU(s) 980 may then share the same virtual address space.
In one implementation of the IOMMU 964, the IOMMU 964 supports virtualization. In this case, it may manage a first set of page tables to map guest/graphics virtual addresses to guest/graphics physical addresses and a second set of page tables to map the guest/graphics physical addresses to system/host physical addresses (e.g., within system memory 966). The base addresses of each of the first and second sets of page tables may be stored in control registers and swapped out on a context switch (e.g., so that the new context is provided with access to the relevant set of page tables). While not illustrated in
The CPU(s) 961, GPUs 980, and I/O devices 962 may be integrated on a single semiconductor chip and/or chip package. The illustrated memory 966 may be integrated on the same chip or may be coupled to the memory controllers 967 via an off-chip interface. In one implementation, the memory 966 comprises GDDR6 memory which shares the same virtual address space as other physical system-level memories, although the underlying principles described herein are not limited to this specific implementation.
The tensor cores 971 may include a plurality of execution units specifically designed to perform matrix operations, which are the fundamental compute operation used to perform deep learning operations. For example, simultaneous matrix multiplication operations may be used for neural network training and inferencing. The tensor cores 971 may perform matrix processing using a variety of operand precisions including single precision floating-point (e.g., 32 bits), half-precision floating point (e.g., 16 bits), integer words (16 bits), bytes (8 bits), and half-bytes (4 bits). For example, a neural network implementation extracts features of each rendered scene, potentially combining details from multiple frames, to construct a high-quality final image.
In deep learning implementations, parallel matrix multiplication work may be scheduled for execution on the tensor cores 971. The training of neural networks, in particular, requires a significant number of matrix dot product operations. In order to process an inner-product formulation of an N×N×N matrix multiply, the tensor cores 971 may include at least N dot-product processing elements. Before the matrix multiply begins, one entire matrix is loaded into tile registers and at least one column of a second matrix is loaded each cycle for N cycles. Each cycle, there are N dot products that are processed.
Matrix elements may be stored at different precisions depending on the particular implementation, including 16-bit words, 8-bit bytes (e.g., INT8) and 4-bit half-bytes (e.g., INT4). Different precision modes may be specified for the tensor cores 971 to ensure that the most efficient precision is used for different workloads (e.g., such as inferencing workloads which can tolerate quantization to bytes and half-bytes). Supported formats additionally include 64-bit floating point (FP64) and non-IEEE floating point formats such as the bfloat16 format (e.g., Brain floating point), a 16-bit floating point format with one sign bit, eight exponent bits, and eight significand bits, of which seven are explicitly stored. One example includes support for a reduced precision tensor-float (TF32) mode, which performs computations using the range of FP32 (8-bits) and the precision of FP16 (10-bits). Reduced precision TF32 operations can be performed on FP32 inputs and produce FP32 outputs at higher performance relative to FP32 and increased precision relative to FP16. In some examples, one or more 8-bit floating point formats (FP8) are supported.
In some examples the tensor cores 971 support a sparse mode of operation for matrices in which the vast majority of values are zero. The tensor cores 971 include support for sparse input matrices that are encoded in a sparse matrix representation (e.g., coordinate list encoding (COO), compressed sparse row (CSR), compress sparse column (CSC), etc.). The tensor cores 971 also include support for compressed sparse matrix representations in the event that the sparse matrix representation may be further compressed. Compressed, encoded, and/or compressed and encoded matrix data, along with associated compression and/or encoding metadata, can be read by the tensor cores 971 and the non-zero values can be extracted. For example, for a given input matrix A, a non-zero value can be loaded from the compressed and/or encoded representation of at least a portion of matrix A. Based on the location in matrix A for the non-zero value, which may be determined from index or coordinate metadata associated with the non-zero value, a corresponding value in input matrix B may be loaded. Depending on the operation to be performed (e.g., multiply), the load of the value from input matrix B may be bypassed if the corresponding value is a zero value. In some examples, the pairings of values for certain operations, such as multiply operations, may be pre-scanned by scheduler logic and only operations between non-zero inputs are scheduled. Depending on the dimensions of matrix A and matrix B and the operation to be performed, output matrix C may be dense or sparse. Where output matrix C is sparse and depending on the configuration of the tensor cores 971, output matrix C may be output in a compressed format, a sparse encoding, or a compressed sparse encoding.
The ray tracing cores 972 may accelerate ray tracing operations for both real-time ray tracing and non-real-time ray tracing implementations. In particular, the ray tracing cores 972 may include ray traversal/intersection circuitry for performing ray traversal using bounding volume hierarchies (BVHs) and identifying intersections between rays and primitives enclosed within the BVH volumes. The ray tracing cores 972 may also include circuitry for performing depth testing and culling (e.g., using a Z buffer or similar arrangement). In one implementation, the ray tracing cores 972 perform traversal and intersection operations in concert with the image denoising techniques described herein, at least a portion of which may be executed on the tensor cores 971. For example, the tensor cores 971 may implement a deep learning neural network to perform denoising of frames generated by the ray tracing cores 972. However, the CPU(s) 961, graphics cores 970, and/or ray tracing cores 972 may also implement all or a portion of the denoising and/or deep learning algorithms.
In addition, as described above, a distributed approach to denoising may be employed in which the GPU 980 is in a computing device coupled to other computing devices over a network or high-speed interconnect. In this distributed approach, the interconnected computing devices may share neural network learning/training data to improve the speed with which the overall system learns to perform denoising for different types of image frames and/or different graphics applications.
The ray tracing cores 972 may process all BVH traversal and/or ray-primitive intersections, saving the graphics cores 970 from being overloaded with thousands of instructions per ray. For example, each ray tracing core 972 includes a first set of specialized circuitry for performing bounding box tests (e.g., for traversal operations) and/or a second set of specialized circuitry for performing the ray-triangle intersection tests (e.g., intersecting rays which have been traversed). Thus, for example, the multi-core group 965A can simply launch a ray probe, and the ray tracing cores 972 independently perform ray traversal and intersection and return hit data (e.g., a hit, no hit, multiple hits, etc.) to the thread context. The other cores 970, 971 are freed to perform other graphics or compute work while the ray tracing cores 972 perform the traversal and intersection operations.
Optionally, each ray tracing core 972 may include a traversal unit to perform BVH testing operations and/or an intersection unit which performs ray-primitive intersection tests. The intersection unit generates a “hit”, “no hit”, or “multiple hit” response, which it provides to the appropriate thread. During the traversal and intersection operations, the execution resources of the other cores (e.g., graphics cores 970 and tensor cores 971) are freed to perform other forms of graphics work.
In some examples described below, a hybrid rasterization/ray tracing approach is used in which work is distributed between the graphics cores 970 and ray tracing cores 972.
The ray tracing cores 972 (and/or other cores 970, 971) may include hardware support for a ray tracing instruction set such as Microsoft's DirectX Ray Tracing (DXR) which includes a DispatchRays command, as well as ray-generation, closest-hit, any-hit, and miss shaders, which enable the assignment of unique sets of shaders and textures for each object. Another ray tracing platform which may be supported by the ray tracing cores 972, graphics cores 970 and tensor cores 971 is Vulkan API (e.g., Vulkan version 1.1.85 and later). Note, however, that the underlying principles described herein are not limited to any particular ray tracing ISA.
In general, the various cores 972, 971, 970 may support a ray tracing instruction set that includes instructions/functions for one or more of ray generation, closest hit, any hit, ray-primitive intersection, per-primitive and hierarchical bounding box construction, miss, visit, and exceptions. More specifically, some examples include ray tracing instructions to perform one or more of the following functions:
In some examples the ray tracing cores 972 may be adapted to accelerate general-purpose compute operations that can be accelerated using computational techniques that are analogous to ray intersection tests. A compute framework can be provided that enables shader programs to be compiled into low level instructions and/or primitives that perform general-purpose compute operations via the ray tracing cores. Exemplary computational problems that can benefit from compute operations performed on the ray tracing cores 972 include computations involving beam, wave, ray, or particle propagation within a coordinate space. Interactions associated with that propagation can be computed relative to a geometry or mesh within the coordinate space. For example, computations associated with electromagnetic signal propagation through an environment can be accelerated via the use of instructions or primitives that are executed via the ray tracing cores. Diffraction and reflection of the signals by objects in the environment can be computed as direct ray-tracing analogies.
Ray tracing cores 972 can also be used to perform computations that are not directly analogous to ray tracing. For example, mesh projection, mesh refinement, and volume sampling computations can be accelerated using the ray tracing cores 972. Generic coordinate space calculations, such as nearest neighbor calculations can also be performed. For example, the set of points near a given point can be discovered by defining a bounding box in the coordinate space around the point. BVH and ray probe logic within the ray tracing cores 972 can then be used to determine the set of point intersections within the bounding box. The intersections constitute the origin point and the nearest neighbors to that origin point. Computations that are performed using the ray tracing cores 972 can be performed in parallel with computations performed on the graphics cores 972 and tensor cores 971. A shader compiler can be configured to compile a compute shader or other general-purpose graphics processing program into low level primitives that can be parallelized across the graphics cores 970, tensor cores 971, and ray tracing cores 972.
Building larger and larger silicon dies is challenging for a variety of reasons. As silicon dies become larger, manufacturing yields become smaller and process technology requirements for different components may diverge. On the other hand, in order to have a high-performance system, key components should be interconnected by high speed, high bandwidth, low latency interfaces. These contradicting needs pose a challenge to high performance chip development.
Embodiments described herein provide techniques to disaggregate an architecture of a system on a chip integrated circuit into multiple distinct chiplets that can be packaged onto a common chassis. In some examples, a graphics processing unit or parallel processor is composed from diverse silicon chiplets that are separately manufactured. A chiplet is an at least partially packaged integrated circuit that includes distinct units of logic that can be assembled with other chiplets into a larger package. A diverse set of chiplets with different IP core logic can be assembled into a single device. Additionally the chiplets can be integrated into a base die or base chiplet using active interposer technology. The concepts described herein enable the interconnection and communication between the different forms of IP within the GPU. The development of IPs on different process may be mixed. This avoids the complexity of converging multiple IPs, especially on a large SoC with several flavors IPs, to the same process.
Enabling the use of multiple process technologies improves the time to market and provides a cost-effective way to create multiple product SKUs. For customers, this means getting products that are more tailored to their requirements in a cost effective and timely manner. Additionally, the disaggregated IPs are more amenable to being power gated independently, components that are not in use on a given workload can be powered off, reducing overall power consumption.
As shown in
The various chiplets can be bonded to a base die 1110 and configured to communicate with each other and logic within the base die 1110 via an interconnect layer 1112. In some examples, the base die 1110 can include global logic 1101, which can include scheduler 1111 and power management 1121 logic units, an interface 1102, a dispatch unit 1103, and an interconnect fabric module 1108 coupled with or integrated with one or more L3 cache banks 1109A-1109N. The interconnect fabric 1108 can be an inter-chiplet fabric that is integrated into the base die 1110. Logic chiplets can use the fabric 1108 to relay messages between the various chiplets. Additionally, L3 cache banks 1109A-1109N in the base die and/or L3 cache banks within the memory chiplets 1106 can cache data read from and transmitted to DRAM chiplets within the memory chiplets 1106 and to system memory of a host.
In some examples the global logic 1101 is a microcontroller that can execute firmware to perform scheduler 1111 and power management 1121 functionality for the parallel processor 1120. The microcontroller that executes the global logic can be tailored for the target use case of the parallel processor 1120. The scheduler 1111 can perform global scheduling operations for the parallel processor 1120. The power management 1121 functionality can be used to enable or disable individual chiplets within the parallel processor when those chiplets are not in use.
The various chiplets of the parallel processor 1120 can be designed to perform specific functionality that, in existing designs, would be integrated into a single die. A set of compute chiplets 1105 can include clusters of compute units (e.g., execution units, streaming multiprocessors, etc.) that include programmable logic to execute compute or graphics shader instructions. A media chiplet 1104 can include hardware logic to accelerate media encode and decode operations. Memory chiplets 1106 can include volatile memory (e.g., DRAM) and one or more SRAM cache memory banks (e.g., L3 banks).
As shown in
At least a portion of the components within the illustrated chiplet 1130 can also be included within logic embedded within the base die 1110 of
Thus, while various examples described herein use the term SOC to describe a device or system having a processor and associated circuitry (e.g., Input/Output (“I/O”) circuitry, power delivery circuitry, memory circuitry, etc.) integrated monolithically into a single Integrated Circuit (“IC”) die, or chip, the present disclosure is not limited in that respect. For example, in various examples of the present disclosure, a device or system can have one or more processors (e.g., one or more processor cores) and associated circuitry (e.g., Input/Output (“I/O”) circuitry, power delivery circuitry, etc.) arranged in a disaggregated collection of discrete dies, tiles and/or chiplets (e.g., one or more discrete processor core die arranged adjacent to one or more other die such as memory die, I/O die, etc.). In such disaggregated devices and systems the various dies, tiles and/or chiplets can be physically and electrically coupled together by a package structure including, for example, various packaging substrates, interposers, active interposers, photonic interposers, interconnect bridges and the like. The disaggregated collection of discrete dies, tiles, and/or chiplets can also be part of a System-on-Package (“SoP”).
In some examples, graphics processor 1200 includes a geometry pipeline 1220, a media pipeline 1230, a display engine 1240, thread execution logic 1250, and a render output pipeline 1270. In some examples, graphics processor 1200 is a graphics processor within a multi-core processing system that includes one or more general-purpose processing cores. The graphics processor is controlled by register writes to one or more control registers (not shown) or via commands issued to graphics processor 1200 via a ring interconnect 1202. In some examples, ring interconnect 1202 couples graphics processor 1200 to other processing components, such as other graphics processors or general-purpose processors. Commands from ring interconnect 1202 are interpreted by a command streamer 1203, which supplies instructions to individual components of the geometry pipeline 1220 or the media pipeline 1230.
In some examples, command streamer 1203 directs the operation of a vertex fetcher 1205 that reads vertex data from memory and executes vertex-processing commands provided by command streamer 1203. In some examples, vertex fetcher 1205 provides vertex data to a vertex shader 1207, which performs coordinate space transformation and lighting operations to each vertex. In some examples, vertex fetcher 1205 and vertex shader 1207 execute vertex-processing instructions by dispatching execution threads to execution units 1252A-1252B via a thread dispatcher 1231.
In some examples, execution units 1252A-1252B are an array of vector processors having an instruction set for performing graphics and media operations. In some examples, execution units 1252A-1252B have an attached L1 cache 1251 that is specific for each array or shared between the arrays. The cache can be configured as a data cache, an instruction cache, or a single cache that is partitioned to contain data and instructions in different partitions.
In some examples, geometry pipeline 1220 includes tessellation components to perform hardware-accelerated tessellation of 3D objects. In some examples, a programmable hull shader 1211 configures the tessellation operations. A programmable domain shader 1217 provides back-end evaluation of tessellation output. A tessellator 1213 operates at the direction of hull shader 1211 and contains special purpose logic to generate a set of detailed geometric objects based on a coarse geometric model that is provided as input to geometry pipeline 1220. In some examples, if tessellation is not used, tessellation components (e.g., hull shader 1211, tessellator 1213, and domain shader 1217) can be bypassed.
In some examples, complete geometric objects can be processed by a geometry shader 1219 via one or more threads dispatched to execution units 1252A-1252B, or can proceed directly to the clipper 1229. In some examples, the geometry shader operates on entire geometric objects, rather than vertices or patches of vertices as in previous stages of the graphics pipeline. If the tessellation is disabled, the geometry shader 1219 receives input from the vertex shader 1207. In some examples, geometry shader 1219 is programmable by a geometry shader program to perform geometry tessellation if the tessellation units are disabled.
Before rasterization, a clipper 1229 processes vertex data. The clipper 1229 may be a fixed function clipper or a programmable clipper having clipping and geometry shader functions. In some examples, a rasterizer and depth test component 1273 in the render output pipeline 1270 dispatches pixel shaders to convert the geometric objects into per pixel representations. In some examples, pixel shader logic is included in thread execution logic 1250. In some examples, an application can bypass the rasterizer and depth test component 1273 and access un-rasterized vertex data via a stream out unit 1223.
The graphics processor 1200 has an interconnect bus, interconnect fabric, or some other interconnect mechanism that allows data and message passing amongst the major components of the processor. In some examples, execution units 1252A-1252B and associated logic units (e.g., L1 cache 1251, sampler 1254, texture cache 1258, etc.) interconnect via a data port 1256 to perform memory access and communicate with render output pipeline components of the processor. In some examples, sampler 1254, caches 1251, 1258 and execution units 1252A-1252B each have separate memory access paths. In some examples the texture cache 1258 can also be configured as a sampler cache.
In some examples, render output pipeline 1270 contains a rasterizer and depth test component 1273 that converts vertex-based objects into an associated pixel-based representation. In some examples, the rasterizer logic includes a windower/masker unit to perform fixed function triangle and line rasterization. An associated render cache 1278 and depth cache 1279 are also available in some examples. A pixel operations component 1277 performs pixel-based operations on the data, though in some instances, pixel operations associated with 2D operations (e.g., bit block image transfers with blending) are performed by the 2D engine 1241, or substituted at display time by the display controller 1243 using overlay display planes. In some examples, a shared L3 cache 1275 is available to all graphics components, allowing the sharing of data without the use of main system memory.
In some examples, graphics processor media pipeline 1230 includes a media engine 1237 and a video front-end 1234. In some examples, video front-end 1234 receives pipeline commands from the command streamer 1203. In some examples, media pipeline 1230 includes a separate command streamer. In some examples, video front-end 1234 processes media commands before sending the command to the media engine 1237. In some examples, media engine 1237 includes thread spawning functionality to spawn threads for dispatch to thread execution logic 1250 via thread dispatcher 1231.
In some examples, graphics processor 1200 includes a display engine 1240. In some examples, display engine 1240 is external to processor 1200 and couples with the graphics processor via the ring interconnect 1202, or some other interconnect bus or fabric. In some examples, display engine 1240 includes a 2D engine 1241 and a display controller 1243. In some examples, display engine 1240 contains special purpose logic capable of operating independently of the 3D pipeline. In some examples, display controller 1243 couples with a display device (not shown), which may be a system integrated display device, as in a laptop computer, or an external display device attached via a display device connector.
In some examples, the geometry pipeline 1220 and media pipeline 1230 are configurable to perform operations based on multiple graphics and media programming interfaces and are not specific to any one application programming interface (API). In some examples, driver software for the graphics processor translates API calls that are specific to a particular graphics or media library into commands that can be processed by the graphics processor. In some examples, support is provided for the Open Graphics Library (OpenGL), Open Computing Language (OpenCL), and/or Vulkan graphics and compute API, all from the Khronos Group. In some examples, support may also be provided for the Direct3D library from the Microsoft Corporation. In some examples, a combination of these libraries may be supported. Support may also be provided for the Open Source Computer Vision Library (OpenCV). A future API with a compatible 3D pipeline would also be supported if a mapping can be made from the pipeline of the future API to the pipeline of the graphics processor.
Program code may be applied to input information to perform the functions described herein and generate output information. The output information may be applied to one or more output devices, in known fashion. For purposes of this application, a processing system includes any system that has a processor, such as, for example, a digital signal processor (DSP), a microcontroller, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a microprocessor, or any combination thereof.
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.
Examples of the mechanisms disclosed herein may be implemented in hardware, software, firmware, or a combination of such implementation approaches. Examples 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.
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 rewritables (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, examples 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 examples may also be referred to as program products.
Although some embodiments have been described in reference to particular implementations, other implementations are possible according to some embodiments. Additionally, the arrangement and/or order of elements or other features illustrated in the drawings and/or described herein need not be arranged in the particular way illustrated and described. Many other arrangements are possible according to some embodiments.
In each system shown in a figure, the elements in some cases may each have a same reference number or a different reference number to suggest that the elements represented could be different and/or similar. However, an element may be flexible enough to have different implementations and work with some or all of the systems shown or described herein. The various elements shown in the figures may be the same or different. Which one is referred to as a first element and which is called a second element is arbitrary.
An embodiment is an implementation or example of the disclosure. Reference in the specification to “an embodiment,” “one embodiment,” “some embodiments,” or “other embodiments” means that a particular feature, structure, or characteristic described in connection with the embodiments is included in at least some embodiments, but not necessarily all embodiments, of the disclosure. The various appearances “an embodiment,” “one embodiment,” or “some embodiments” are not necessarily all referring to the same embodiments.
Not all components, features, structures, characteristics, etc. described and illustrated herein need to be included in a particular embodiment or embodiments. If the specification states a component, feature, structure, or characteristic “may”, “might”, “can”, or “could” be included, for example, that particular component, feature, structure, or characteristic is not required to be included. If the specification or claim refers to “a” or “an” element, that does not mean there is only one of the elements. If the specification or claims refer to “an additional” element, that does not preclude there being more than one of the additional elements.
The above description of illustrated embodiments of the disclosure, including what is described in the Abstract, is not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. While specific embodiments of, and examples for, the disclosure are described herein for illustrative purposes, various equivalent modifications are possible within the scope of the disclosure, as those skilled in the relevant art will recognize.
These modifications can be made to the disclosure in light of the above detailed description. The terms used in the following claims should not be construed to limit the disclosure to the specific embodiments disclosed in the specification and the drawings. Rather, the scope of the disclosure is to be determined entirely by the following claims, which are to be construed in accordance with established doctrines of claim interpretation.
Example 1 provides an exemplary method comprising: responsive to a virtual machine entry to a virtual processing unit of a processor, writing (402) a first count value and a second count value corresponding to the virtual processing unit to a first counter and a second counter of the processor, respectively; incrementing (404) the first counter counts based on the first count value at a fixed frequency and incrementing (404) the second counter counts based on the second count value in proportion of architectural event occurrence of the virtual processing unit; and responsive to a virtual machine exit from the virtual processing unit, reading (406) a resulting first count value and a resulting second count value from the first and second counters respectively to determine a measure of the performance of the virtual processing unit within a corresponding duration between the virtual machine entry and the virtual machine exit.
Example 2 includes the substance of Example 1, wherein the resulting first and second count values indicate the performance of the virtual processing unit at a corresponding performance state of the processor that maps to one or more frequency and voltage settings.
Example 3 includes the substance of Examples 1 to 2, wherein the first and second count values are obtained based on a first value and a second value stored through a virtual machine control structure corresponding to the virtual processing unit.
Example 4 includes the substance of Examples 1 to 3, wherein the first and second values are stored through a virtual machine execution control field of the virtual machine control structure.
Example 5 includes the substance of Examples 1 to 4, wherein one or more of the resulting first and second count values are scaled based on a time-stamp counter (TSC) multiplier or count offset of the virtual processing unit prior to storing as updates of the first and second values through the virtual machine execution control field.
Example 6 includes the substance of Examples 1 to 5, wherein an access request to the first counter or the second counter of the processor is redirected to a corresponding value of the first and second values stored in the virtual machine control structure.
Example 7 includes the substance of Examples 1 to 6, wherein the measure of performance of the virtual processing unit is determined based on a ratio between the resulting first and second count values from the first and second counters of the processor.
Example 8 includes the substance of Examples 1 to 7, wherein the first and second count values corresponding to the virtual processing unit are written to a first register and a second register corresponding to the first and second counters, respectively, and the resulting first and second count values are read from the first and second registers, respectively.
Example 9 includes the substance of Examples 1 to 8, wherein a set of bits corresponding to the first and second registers is set to enable access to the first and second registers prior to accessing the first and second registers in a non-root mode.
Example 10 includes the substance of Examples 1 to 9, wherein the first and second registers are separated from registers to measure performance of the processor in both root mode and non-root mode.
Example 11 provides an exemplary processor (150) comprising: a set of cores (152, 154) to provide one or more virtual processing units; and circuitry (190) coupled to the set of cores to measure performance of the one or more virtual processing units, the circuitry to perform: responsive to a virtual machine entry to a virtual processing unit of a processor, writing (402) a first count value and a second count value corresponding to the virtual processing unit to a first counter and a second counter of the processor, respectively; incrementing (404) the first counter counts based on the first count value at a fixed frequency and incrementing (404) the second counter counts based on the second count value in proportion of architectural event occurrence of the virtual processing unit; and responsive to a virtual machine exit from the virtual processing unit, reading (406) a resulting first count value and a resulting second count value from the first and second counters respectively to determine a measure of the performance of the virtual processing unit within a corresponding duration between the virtual machine entry and the virtual machine exit.
Example 12 includes the substance of Example 11, wherein the resulting first and second count values indicate the performance of the virtual processing unit at a corresponding performance state of the processor that maps to one or more frequency and voltage settings.
Example 13 includes the substance of Examples 11 to 12, wherein the first and second count values are obtained based on a first value and a second value stored through a virtual machine control structure corresponding to the virtual processing unit.
Example 14 includes the substance of Examples 11 to 13, wherein the first and second values are stored through a virtual machine execution control field of the virtual machine control structure.
Example 15 includes the substance of Examples 11 to 14, wherein one or more of the resulting first and second count values are scaled based on a time-stamp counter (TSC) multiplier or count offset of the virtual processing unit prior to storing as updates of the first and second values through the virtual machine execution control field.
Example 16 includes the substance of Examples 11 to 15, wherein an access request to the first counter or the second counter of the processor is redirected to a corresponding value of the first and second values stored in the virtual machine control structure.
Example 17 includes the substance of Examples 11 to 16, wherein the performance of the virtual processing unit is determined based on a ratio between the resulting first and second count values from the first and second counters of the processor.
Example 18 includes the substance of Examples 11 to 17, wherein the first and second count values corresponding to the virtual processing unit are written to a first register and a second register corresponding to the first and second counters, respectively, and the resulting first and second count values are read from the first and second registers, respectively.
Example 19 includes the substance of Examples 11 to 18, wherein a set of bits corresponding to the first and second registers is set to enable access to the first and second registers prior to accessing the first and second registers in a non-root mode.
Example 20 includes the substance of Examples 11 to 19, wherein the first and second registers are separated from registers to measure performance of the processor in both root mode and non-root mode.
Example 21 provides an exemplary computer-readable storage medium storing instructions that when executed by a processor, are capable of causing the processor to perform any of Examples 1 to 10.
Embodiments of the disclosure may include various steps, which have been described above. The steps may be embodied in machine-executable instructions which may be used to cause a general-purpose or special-purpose processor to perform the steps. Alternatively, these steps may be performed by specific hardware components that contain hardwired logic for performing the steps, or by any combination of programmed computer components and custom hardware components.
As described herein, instructions may refer to specific configurations of hardware such as application specific integrated circuits (ASICs) configured to perform certain operations or having a predetermined functionality or software instructions stored in memory embodied in a non-transitory computer-readable medium. Thus, the techniques shown in the Figures can be implemented using code and data stored and executed on one or more electronic devices (e.g., an end station, a network element, etc.). Such electronic devices store and communicate (internally and/or with other electronic devices over a network) code and data using computer machine-readable media, such as non-transitory computer machine-readable storage media (e.g., magnetic disks; optical disks; random access memory; read only memory; flash memory devices; phase-change memory) and transitory computer machine-readable communication media (e.g., electrical, optical, acoustical, or other form of propagated signals-such as carrier waves, infrared signals, digital signals, etc.). In addition, such electronic devices typically include a set of one or more processors coupled to one or more other components, such as one or more storage devices (non-transitory machine-readable storage media), user input/output devices (e.g., a keyboard, a touchscreen, and/or a display), and network connections. The coupling of the set of processors and other components is typically through one or more busses and bridges (also termed as bus controllers). The storage device and signals carrying the network traffic respectively represent one or more machine-readable storage media and machine-readable communication media. Thus, the storage device of a given electronic device typically stores code and/or data for execution on the set of one or more processors of that electronic device. Of course, one or more parts of an embodiment of the disclosure may be implemented using different combinations of software, firmware, and/or hardware. Throughout this detailed description, for the purposes of explanation, numerous specific details were set forth in order to provide a thorough understanding of the present disclosure. It will be apparent, however, to one skilled in the art that the disclosure may be practiced without some of these specific details. In certain instances, well-known structures and functions were not described in elaborate detail in order to avoid obscuring the subject matter of the present disclosure. Accordingly, the scope and spirit of the disclosure should be judged in terms of the claims which follow.
This application is a continuation of International Application No. PCT/CN2023/138063, filed Dec. 12, 2023, which is hereby incorporated by reference.
Number | Date | Country | |
---|---|---|---|
Parent | PCT/CN2023/138063 | Dec 2023 | WO |
Child | 18944849 | US |