SMT/ECO MODE BASED ON CACHE MISS RATE

• US 20110302372A1
• Filed: 06/03/2010
• Published: 12/08/2011
• Est. Priority Date: 06/03/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Poppe, Monty C., Schopp, Joel H., Strosaker, Michael T., Fontenot, Nathan D, Grimm, Ryan P.
• Granted Patent: US 8,285,950 B2
• Time in Patent Office: Days
• US Class Current: 711/122
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 11/348 Circuit details, i.e. tracer hardware
• CPC Class Codes: G06F 2201/81 Threshold
• CPC Class Codes: G06F 2201/88 Monitoring involving counting
• CPC Class Codes: G06F 2201/885 Monitoring specific for caches
• CPC Class Codes: G06F 2209/509 Offload
• CPC Class Codes: G06F 9/3851 from multiple instruction streams, e.g. multistreaming
• CPC Class Codes: G06F 9/3877 using a slave processor, e.g. coprocessor peripheral processor G06F13/12; vector processor G06F15/8053
• CPC Class Codes: G06F 9/5088 involving task migration

Abstract

A computer implemented method for managing an execution mode for a parallel processor is provided. A monitor identifies a first efficiency rate for a first contested resource of the parallel processor operating in a first operating mode. Responsive to identifying the first efficiency rate for the first contested resource, the monitor identifies whether the first efficiency rate for the contested resource of the parallel processor operating in the first operating mode exceeds a threshold. Responsive to identifying that the efficiency rate for the contested resource exceeds the threshold, an operation of the parallel processor is changed to a second operating mode.

Claims

1. A computer implemented method for managing an execution mode for a parallel processor, the method comprising:

identifying a first efficiency rate for a first contested resource of the parallel processor operating in a first operating mode;

responsive to identifying the first efficiency rate for the first contested resource, identifying whether the first efficiency rate for the contested resource of the parallel processor operating in the first operating mode exceeds a threshold; and

responsive to identifying that the efficiency rate for the contested resource exceeds the threshold, changing an operation of the parallel processor to a second operating mode.

2. The computer implemented method of claim 1 further comprising:

identifying a second efficiency rate for the first contested resource of the parallel processor operating in the second operating mode;

responsive to identifying the second efficiency rate for the first contested resource of the parallel processor operating in the second operating mode, identifying whether the second efficiency rate exceeds the first efficiency rate; and

responsive to identifying that the second efficiency rate exceeds the first efficiency rate, changing the operation of the parallel processor to the first operating mode.

3. The computer implemented method of claim 2, wherein the step of identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a simultaneous multi-threaded processor.

4. The computer implemented method of claim 3, wherein the step of identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a cache allocated to a plurality of threads of the simultaneous multi-threaded processor, wherein the first efficiency rate is a cache miss rate of an L2 cache and an L3 cache of the simultaneous multi-threaded processor.

5. The computer implemented method of claim 2, wherein the step of identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a multi core processor.

6. The computer implemented method of claim 5, wherein the step of identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a cache allocated to a plurality of cores of the multi core processor, wherein the first efficiency rate is a cache miss rate of an L2 cache and an L3 cache of the multi core processor.

7. The computer implemented method of claim 5, wherein the step of identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a coprocessor, wherein the first efficiency rate is selected from the group consisting of a time that an off-loaded instruction remains in a queue of the coprocessor, a time of execution of the off-loaded instruction by the coprocessor, and a time that a core of the multi core processor is idle.

8. A computer-readable medium having a computer program product encoded thereon for managing an execution mode for a parallel processor, the computer program product comprising:

instructions for identifying a first efficiency rate for a first contested resource of the parallel processor operating in a first operating mode;

instructions, responsive to identifying the first efficiency rate for the first contested resource, for identifying whether the first efficiency rate for the contested resource of the parallel processor operating in the first operating mode exceeds a threshold; and

instructions, responsive to identifying that the efficiency rate for the contested resource exceeds the threshold, for changing an operation of the parallel processor to a second operating mode.

9. The computer-readable medium of claim 8, wherein the computer program product further comprises:

instructions for identifying a second efficiency rate for the first contested resource of the parallel processor operating in a second operating mode;

instructions, responsive to identifying the second efficiency rate for the first contested resource of the parallel processor operating in the second operating mode, for identifying whether the second efficiency rate exceeds the first efficiency rate; and

instructions, responsive to identifying that the second efficiency rate exceeds the first efficiency rate, for changing the operation of the parallel processor to the first operating mode.

10. The computer-readable medium of claim 9, wherein the instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a simultaneous multi-threaded processor.

11. The computer-readable medium of claim 10, wherein the instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a cache allocated to a plurality of threads of the simultaneous multi-threaded processor, wherein the first efficiency rate is a cache miss rate of an L2 cache and an L3 cache of the simultaneous multi-threaded processor.

12. The computer-readable medium of claim 9, wherein the instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a multi core processor.

13. The computer-readable medium of claim 12, wherein the instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is the cache allocated to a plurality of cores of the multi core processor, wherein the first efficiency rate is a cache miss rate of an L2 cache and an L3 cache of the multi core processor.

14. The computer-readable medium of claim 12, wherein the instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises:

instructions for identifying the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a coprocessor, wherein the first efficiency rate is selected from the group consisting of a time that an off-loaded instruction remains in a queue of the coprocessor, a time of execution of the off-loaded instruction by the coprocessor, and a time that a core of the multi core processor is idle.

15. A data processing system comprising:

a storage having a computer program product encoded thereon for managing an execution mode for a parallel processor;

a bus connecting the storage to a processor; and

a processor, wherein the processor executes the computer program product;

to identify a first efficiency rate for a first contested resource of the parallel processor operating in a first operating mode;

responsive to identifying the first efficiency rate for the first contested resource, to identify whether the first efficiency rate for the contested resource of the parallel processor operating in the first operating mode exceeds a threshold; and

responsive to identifying that the efficiency rate for the contested resource exceeds the threshold, to change an operation of the parallel processor to a second operating mode.

16. The data processing system of claim 15, wherein the processor further executes the computer program product:

to identify a second efficiency rate for the first contested resource of the parallel processor operating in a second operating mode;

responsive to identifying the second efficiency rate for the first contested resource of the parallel processor operating in the second operating mode, to identify whether the second efficiency rate exceeds the first efficiency rate; and

responsive to identifying that the second efficiency rate exceeds the first efficiency rate, to change the operation of the parallel processor to the first operating mode.

17. The data processing system of claim 16, wherein the processor executing the computer program product to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises the processor executing the computer program product:

to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a simultaneous multi-threaded processor.

18. The data processing system of claim 17, wherein the processor executing the computer program product to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises the processor executing the computer program product:

to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is a cache allocated to a plurality of threads of the simultaneous multi-threaded processor, wherein the first efficiency rate is a cache miss rate of an L2 cache and an L3 cache of the simultaneous multi-threaded processor.

19. The data processing system of claim 16, wherein the processor executing the computer program product to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises the processor executing the computer program product:

to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the parallel processor is a multi core processor.

20. The data processing system of claim 19, wherein the processor executing the computer program product to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises the processor executing the computer program product:

to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is the cache allocated to the plurality of cores of the multi core processor, wherein the first efficiency rate is the cache miss rate of an L2 cache and an L3 cache of the multi core processor.

21. The data processing system of claim 19, wherein the processor executing the computer program product to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode further comprises the processor executing the computer program product:

to identify the first efficiency rate for the first contested resource of the parallel processor operating in the first operating mode, wherein the contested resource is the coprocessor, wherein the first efficiency rate is selected from the group consisting of a time that an off-loaded instruction remains in a queue of the coprocessor, a time of execution of the off-loaded instruction by the coprocessor, and a time that the core of the multi core processor is idle.