Power efficient processor architecture

Patent metadata

Field	Value
Publication number	US-9360927-B2
Application number	US-201113992361-A
Country	US
Kind code	B2
Filing date	Sep 6, 2011
Priority date	Sep 6, 2011
Publication date	Jun 7, 2016
Grant date	Jun 7, 2016

Abstract

In one embodiment, the present invention includes a method for receiving an interrupt from an accelerator, sending a resume signal directly to a small core responsive to the interrupt and providing a subset of an execution state of the large core to the first small core, and determining whether the small core can handle a request associated with the interrupt, and performing an operation corresponding to the request in the small core if the determination is in the affirmative, and otherwise providing the large core execution state and the resume signal to the large core. Other embodiments are described and claimed.

First claim

What is claimed is:

1. A processor comprising:
a first core to execute instructions;
a second core to execute instructions, the second core heterogeneous to and smaller than the first core; and
a power control unit (PCU) including a logic to cause the second core and not the first core to be woken responsive to an interrupt when the first and second cores are in a low power state, wherein the PCU is to cause the first core and not the second core to be woken responsive to the interrupt when an entry of a table stored in the PCU indicates that the second core incurred an undefined fault responsive to a previous interrupt of the same type as the interrupt.

2. The processor of claim 1 , wherein the logic is to always cause the second core and not the first core to be woken responsive to the interrupt.

3. The processor of claim 1 , wherein the logic is to provide a subset of an execution state of the first core to the second core responsive to the interrupt.

4. The processor of claim 3 , wherein responsive to the determination that the second core cannot handle at least one requested operation associated with the interrupt, the logic is to obtain the subset of the execution state of the first core from the second core and to merge the execution state subset with a remainder of the execution state of the first core stored in a temporary storage area.

5. The processor of claim 1 , wherein the processor comprises a multicore processor, the logic comprising:
a wakeup logic;
a state transfer logic;
an undefined handling logic; and
an interrupt history storage.

6. The processor of claim 1 , further comprising an accelerator coupled to the logic, the accelerator to perform a task and to send the interrupt to the logic upon completion of the task.

7. The processor of claim 6 , wherein the second core is to handle the interrupt when the interrupt comprises a request for a data movement operation.

8. The processor of claim 6 , wherein the second core is to cause the wakeup signal to be sent to the first core to enable the first core to handle the interrupt when the interrupt comprises a request for a vector operation.

9. The processor of claim 8 , wherein the logic is to receive an undefined instruction fault from the second core, determine that the second core cannot handle the vector operation, obtain an execution state from the second core, merge the execution state with at least a portion of an execution state of the first core stored in a temporary storage area, and cause the merged execution state to be sent to the first core.

10. The processor of claim 1 , wherein the logic is to analyze a plurality of interrupts and if a majority of the plurality of interrupts are to be handled by the first core, the logic is to not wake the second core responsive to the interrupt and instead wake the first core.

11. A method comprising:
receiving an interrupt from an accelerator in a logic of a processor including a first small core, a first large core and the accelerator, when the first small core and the first large core are in a low power state;
accessing an entry of a table based on a type of the interrupt and determining whether to send a resume signal directly to the first small core or the first large core based on the entry;
sending the resume signal directly to the first small core responsive to the interrupt and providing a subset of an execution state of the first large core to the first small core; and
determining, in the first small core, whether the first small core can handle a request associated with the interrupt, and if so performing an operation corresponding to the request in the first small core.

12. The method of claim 11 , further comprising if the first small core cannot handle the request, obtaining the execution state subset from the first small core, merging the execution state subset with a stored execution state of the first large core, and sending a wakeup signal and the merged execution state to the first large core.

13. The method of claim 12 , further comprising thereafter performing the operation corresponding to the request in the first large core.

14. The method of claim 11 , further comprising receiving the interrupt with a hint to indicate whether the interrupt should be directed to the first small core or the first large core.

15. A system comprising:
a multicore processor including a first plurality of cores and a second plurality of cores, the second plurality of cores having lower thermal design power than the first plurality of cores, an accelerator, and a power control unit (PCU), wherein the PCU is to receive an interrupt from the accelerator when the first plurality of cores and the second plurality of cores are in a low power state, send a resume signal directly to a first of the second plurality of cores responsive to the interrupt and provide a subset of an execution state of a first of the first plurality of cores to the first of the second plurality of cores, wherein the PCU is to access an entry of a table stored in the PCU using the interrupt to determine whether to send the resume signal to the first one of the first or second plurality of cores, wherein the PCU is to send the resume signal to the first one of the first plurality of cores when the entry indicates that one of the second plurality of cores incurred an undefined fault responsive to a previous interrupt of the same type as the interrupt; and
a dynamic random access memory (DRAM) coupled to the multicore processor.

16. The system of claim 15 , wherein the first plurality of cores are of a heterogeneous design from the second plurality of cores.

17. The system of claim 15 , wherein the second plurality of cores are transparent to an operating system (OS).

Assignees

• Herdrich Andrew J
• Illikkal Rameshkumar G
• Iyer Ravishankar
• Srinivasan Sadogopan
• Moses Jaideep
• Makineni Srihari
• Intel Corp

Inventors

• Herdrich Andrew J
• Illikkal Rameshkumar G
• Iyer Ravishankar
• Srinivasan Sadogopan
• Moses Jaideep
• Makineni Srihari

Classifications

• G06F1/3293 Primary

  by switching to a less power-consuming processor, e.g. sub-CPU · CPC title

• G06F2212/60

  Details of cache memory · CPC title

• G06F9/5094 Primary

  where the allocation takes into account power or heat criteria (power management in computers in general G06F1/3203; thermal management in computers in general G06F1/206) · CPC title

• G06F2212/1028

  Power efficiency · CPC title

• Y02B70/30

  Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level · CPC title