Rotational graphics sub-slice and execution unit power down to improve power performance efficiency

• US 9,703,364 B2
• Filed: 09/29/2012
• Issued: 07/11/2017
• Est. Priority Date: 09/29/2012
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Hurd, Linda L.
• Primary Examiner(s): Rehman, Mohammed
• Assistant Examiner(s): Mazyad, Alyaa T
• Application Number:
  US13/631,913
• Publication Number:
  US 20140095906A1
• Time in Patent Office: 1,746 Days
• Field of Search: 713300, 713320, 713324
• CPC Class Codes: G06F 1/3278 Power saving in modem or I/O interface
• CPC Class Codes: G06F 1/3293 by switching to a less power-consuming processor, e.g. sub-CPU
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

Methods and apparatus relating to rotational graphics sub-slice and Execution Unit (EU) power down to improve power performance efficiency are described. In one embodiment, power-gating is rotated amongst single sub-slices within each slice of a plurality of slices based on an indication to reduce power consumption of a computational logic. The computational logic includes the plurality of slices and each of the plurality of slices includes a plurality of sub-slices to perform one or more computations. Other embodiments are also disclosed and claimed.

Claims

1. A processor comprising:

computational logic having a plurality of slices, wherein each of the plurality of slices is to comprise a plurality of sub-slices to perform one or more computations related to one or more graphics operations; and

logic to cause rotation of power-gating amongst single sub-slices within each slice of the plurality of slices based on an indication to reduce power consumption of the computational logic, wherein the indication to reduce power consumption of the computational logic is to be generated based at least in part on a ratio of leakage power and dynamic power, wherein the one or more graphics operations are to be performed on a plurality of frames, wherein logic to cause rotation of the power-gating is to cause the rotation of the power-gating partway through a frame or at a frame boundary, wherein residency in each rotation of power-gating is capable to differ for each of the sub-slices, wherein logic to cause rotation of power-gating amongst single sub-slices within each slice of the plurality of slices is to cause each powered down single sub-slice within each slice of the plurality of slices to be powered up in response to expiration of a timer.

2. The processor of claim 1, wherein each sub-slice is to comprise one or more execution units.

3. The processor of claim 2, wherein each of the one or more execution units is to comprise a plurality of processing elements.

4. The processor of claim 3, wherein each of the plurality of processing elements is to comprise one or more ALUs (Arithmetic Logic Units), wherein each of the one or more ALUs is to perform the role of a shader processor.

5. The processor of claim 1, wherein logic to cause rotation of power-gating is to cause rotation of power-gating based on information detected at one or more sensors.

6. The processor of claim 5, wherein the one or more sensors are to detect variations in one or more of:

temperature, operating frequency, operating voltage, operating current, dynamic capacitance, power consumption, and inter-core communication activity.

7. The processor of claim 1, wherein the logic to cause rotation of power-gating amongst single sub-slices is to cause a first time period spent in a first rotation to have a different period than a second time period spent in a second rotation.

8. The processor of claim 1, wherein the plurality of frames are to form a scene.

9. The processor of claim 1, wherein one or more of the computational logic, a voltage regulator, the logic to cause rotation of power-gating, or a memory are on a single integrated circuit die.

10. The processor of claim 1, wherein the indication to reduce power consumption of the computational logic is to be generated based at least in part on the ratio of leakage power and dynamic power within an allotted power budget.

11. The processor of claim 1, wherein the residency in each rotation of power-gating is capable to differ from a first workload to a second workload.

12. A method comprising:

rotating power-gating amongst single sub-slices within each slice of a plurality of slices based on an indication to reduce power consumption of a computational logic, wherein the computational logic comprises the plurality of slices and wherein each of the plurality of slices comprises a plurality of sub-slices to perform one or more computations related to one or more graphics operations, wherein the indication to reduce power consumption of the computational logic is generated based at least in part on a ratio of leakage power and dynamic power, wherein the one or more graphics operations are performed on a plurality of frames, wherein the rotation of the power-gating occurs partway through a frame or at a frame boundary, wherein residency in each rotation of power-gating is capable to differ for Title;

ROTATIONAL GRAPHICS SUB-SLICE AND EXECUTION UNIT POWER DOWN TO IMPROVE POWER PERFORMANCE each of the sub-slices, wherein each powered down single sub-slice within each slice of the plurality of slices is caused to power up in response to expiration of a timer.

13. The method of claim 12, further comprising one or more sensors detecting variations in one or more of:

temperature, operating frequency, operating voltage, operating current, dynamic capacitance, power consumption, and inter-core communication activity.

14. The method of claim 12, wherein the residency in each rotation of power-gating is capable to differ from a first workload to a second workload.

15. A system comprising:

a processor having one or more processor cores;

memory to store data to be accessed by at least one of the one or more processor cores;

the processor comprising;

computational logic having a plurality of slices, wherein each of the plurality of slices is to comprise a plurality of sub-slices to perform one or more computations related to one or more graphics operations; and

logic to cause rotation of power-gating amongst single sub-slices within each slice of the plurality of slices based on an indication to reduce power consumption of the computational logic, wherein the indication to reduce power consumption of the computational logic is to be generated based at least in part on a ratio of leakage power and dynamic power, wherein the one or more graphics operations are to be performed on a plurality of frames, wherein logic to cause rotation of the power-gating is to cause the rotation of the power-gating partway through a frame or at a frame boundary, wherein residency in each rotation of power-gating is capable to differ for each of the sub-slices, wherein logic to cause rotation of power-gating amongst single sub-slices within each slice of the plurality of slices is to cause each powered down single sub-slice within each slice of the plurality of slices to be powered up in response to expiration of a timer.

16. The system of claim 15, wherein each sub-slice is to comprise one or more execution units.

17. The system of claim 16, wherein each of the one or more execution units is to comprise a plurality of processing elements.

18. The system of claim 16, further comprising a display device, coupled to the memory, to present at least one frame of a scene.

19. The system of claim 15, wherein logic to cause rotation of power-gating is to cause rotation of power-gating based on information detected at one or more sensors.

20. The system of claim 19, wherein the one or more sensors are to detect variations in one or more of:

temperature, operating frequency, operating voltage, operating current, dynamic capacitance, power consumption, and inter-core communication activity.

21. The system of claim 19, wherein the one or more sensors are to be proximate to the plurality of sub-slices.

22. The system of claim 15, wherein one or more of the computational logic, a voltage regulator, the logic to cause rotation of power-gating, or the memory are on a single integrated circuit die.

23. The system of claim 15, wherein the residency in each rotation of power-gating is capable to differ from a first workload to a second workload.

24. A non-transitory computer-readable medium comprising one or more instructions that when executed on a processor configure the processor to perform one or more operations to:

rotate power-gating amongst single sub-slices within each slice of a plurality of slices based on an indication to reduce power consumption of a computational logic, wherein the computational logic comprises the plurality of slices and wherein each of the plurality of slices comprises a plurality of sub-slices to perform one or more computations related to one or more graphics operations, wherein the indication to reduce power Title;

ROTATIONAL GRAPHICS SUB-SLICE AND EXECUTION UNIT POWER DOWN TO IMPROVE POWER PERFORMANCE consumption of the computational logic is to be generated based at least in part on a ratio of leakage power and dynamic power, wherein the one or more graphics operations are performed on a plurality of frames, wherein the rotation of the power-gating occurs partway through a frame or at a frame boundary, wherein residency in each rotation of power-gating is capable to differ for each of the sub-slices, wherein one or more instructions, when executed on the processor, are to configure the processor to cause each powered down single sub-slice within each slice of the plurality of slices to be powered up in response to expiration of a timer.

25. The non-transitory computer-readable medium of claim 24, further comprising one or more instructions that when executed on the processor configure the processor to rotate power-gating amongst single sub-slices within each slice of the plurality of slices based on information detected at one or more sensors.

26. The non-transitory computer-readable medium of claim 24, wherein the residency in each rotation of power-gating is capable to differ from a first workload to a second workload.