Systems, methods and devices for determining work placement on processor cores

Patent metadata

Field	Value
Publication number	US-12293237-B2
Application number	US-202318545912-A
Country	US
Kind code	B2
Filing date	Dec 19, 2023
Priority date	Jan 15, 2016
Publication date	May 6, 2025
Grant date	May 6, 2025

Abstract

Apparatuses, methods and storage medium for computing including determination of work placement on processor cores are disclosed herein. In embodiments, an apparatus may include one or more processors, devices, and/or circuitry to identify a favored core of the processor cores. The one or more processors, devices, and/or circuitry may be configured to determine whether to migrate a thread to or from the favored core. In some embodiments, the determination may be by a process executed by a driver and/or by an algorithm executed by a power control unit of the processor.

First claim

What is claimed is:

1. A processor, comprising:
a plurality of cores that are architecturally identical;
circuitry to store data indicating physical characteristics of each core of the plurality of cores, the physical characteristics including a maximum frequency of each core of the plurality of cores, at least one core of the plurality of cores having different physical characteristics than another core of the plurality of cores; and
power control circuitry to control frequencies and voltages of the plurality of cores in accordance with corresponding per-core performance values, the power control circuitry including an interface to receive performance requests from system software and to provide information related to the per-core performance values to the system software;
execution circuitry to execute instructions of a driver to determine the per-core performance values,
wherein when a first operational mode is enabled, the execution circuitry is to:
determine a first ranking of the plurality of cores based on the physical characteristics of the plurality of cores, and
provide the first ranking to the system software, and

wherein when the first operational mode is not enabled, the execution circuitry is to:
determine a second ranking of the plurality of cores based on the physical characteristics of the plurality of cores, the second ranking different from the first ranking, and
provide the second ranking to the system software.

2. The processor of claim 1 wherein when the first operational mode is enabled, the power control circuitry is to:
receive a request from the system software indicating a requested performance level for a particular core of the plurality of cores, and
control the frequency and voltage of the particular core based on the requested performance level.

3. The processor of claim 1 , wherein a candidate core of the plurality of cores is to be selected for executing a current workload based on the first ranking when the first operational mode is enabled and based on the second ranking when the first operational mode is not enabled.

4. The processor of claim 3 , wherein the candidate core is to be selected by the power control circuitry.

5. The processor of claim 3 , wherein the candidate core is to be selected by the system software.

6. The processor of claim 1 , wherein the circuitry to store data indicating physical characteristics of each core of the plurality of cores comprises a read-only memory.

7. The processor of claim 1 , wherein the circuitry to store data indicating physical characteristics of each core of the plurality of cores comprises a set of fuses to be programmed to indicate the physical characteristics of each core of the plurality of cores.

8. The processor of claim 1 wherein, when the first operational mode is enabled, the power control circuitry is to receive a request from the system software for a specific level of performance of a first core for executing a current workload.

9. A non-transitory machine-readable medium having program code stored thereon which, when executed by a processor, is to cause the processor to perform:
receiving, over an interface of the processor, performance requests from system software, the performance requests associated with one or more cores of a plurality of cores that are architecturally identical within the processor;
providing information related to per-core performance values to the system software, the per-core performance values based, at least in part, on data indicating physical characteristics of each core of the plurality of cores, the physical characteristics including a maximum frequency of each core of the plurality of cores, at least one core of the plurality of cores having different physical characteristics than another core of the plurality of cores; and
controlling, by power control circuitry of the processor, frequencies and voltages of the plurality of cores in accordance with corresponding per-core performance values;
wherein when a first operational mode is enabled, execution circuitry of the processor performs:
determining a first ranking of the plurality of cores based on the physical characteristics of the plurality of cores, and
providing the first ranking to the system software, and

wherein when the first operational mode is not enabled, the execution circuitry of the processor performs:
determining a second ranking of the plurality of cores based on the physical characteristics of the plurality of cores, the second ranking different from the first ranking, and
providing the second ranking to the system software.

10. The non-transitory machine-readable medium of claim 9 wherein, when the first operational mode is enabled, the performance requests are to indicate a specific level of performance of a first core for executing a current workload.

11. The non-transitory machine-readable medium of claim 9 wherein when the first operational mode is enabled, the program code is to cause the processor to perform:
receiving a request from the system software indicating a requested performance level for a particular core of the plurality of cores, and
controlling the frequency and voltage of the particular core based on the requested performance level.

12. The non-transitory machine-readable medium of claim 9 , wherein a candidate core of the plurality of cores is to be selected for executing a current workload based on the first ranking when the first operational mode is enabled and based on the second ranking when the first operational mode is not enabled.

13. The non-transitory machine-readable medium of claim 12 , wherein the candidate core is to be selected by the system software.

14. The non-transitory machine-readable medium of claim 9 , wherein data indicating physical characteristics of each core of the plurality of cores is stored in a read-only memory.

15. The processor of claim 1 , wherein the power control circuitry further comprises one or more registers to store an indication of a power state of each core of the plurality of cores.

16. The processor of claim 1 , further comprising:
shared circuitry external to the plurality of cores, the shared circuitry including an interconnect and last level cache (LLC) coupled to the plurality of cores, wherein the power control circuitry is to control a shared circuitry frequency and voltage in accordance with a power state of at least a portion of the shared circuitry.

17. The processor of claim 1 , wherein the power control circuitry comprises a programmable state machine or a microcontroller to execute power control code.

18. The processor of claim 1 , wherein the interface includes at least one of: mailbox registers and model-specific registers (MSRs).

Assignees

• Intel Corp

Inventors

• Therien Guy M
• Powell Michael D
• Ramani Venkatesh
• Biswas Arijit
• Sotomayor Guy G

Classifications

• G06F9/3009

  Thread control instructions · CPC title

• G06F1/3296

  by lowering the supply or operating voltage · CPC title

• G06F1/324 Primary

  by lowering clock frequency · CPC title

• Y02D10/00

  Energy efficient computing, e.g. low power processors, power management or thermal management · CPC title

• G06F1/329

  by task scheduling · CPC title