System, apparatus and method for dynamic pipeline stage control of data path dominant circuitry of an integrated circuit

Patent metadata

Field	Value
Publication number	US-11132201-B2
Application number	US-201916725041-A
Country	US
Kind code	B2
Filing date	Dec 23, 2019
Priority date	Dec 23, 2019
Publication date	Sep 28, 2021
Grant date	Sep 28, 2021

Abstract

In an embodiment, a data path circuit includes: a plurality of pipeline stages coupled between an input of the data path circuit and an output of the data path circuit; and a first selection circuit coupled between a first pipeline stage and a second pipeline stage, the first selection circuit having a first input to receive an input to the first pipeline stage and a second input to receive an output of the first pipeline stage and controllable to output one of the input to the first pipeline stage and the output of the first pipeline stage. A bypass controller coupled to the data path circuit may control the first selection circuit based at least in part on an operating frequency of the data path circuit. Other embodiments are described and claimed.

First claim

What is claimed is:

1. An apparatus comprising:
a data path circuit having:
a plurality of pipeline stages coupled between an input of the data path circuit and an output of the data path circuit; and
a first selection circuit coupled between a first pipeline stage and a second pipeline stage, the first selection circuit having a first input to receive an input to the first pipeline stage and a second input to receive an output of the first pipeline stage and controllable to output one of the input to the first pipeline stage and the output of the first pipeline stage; and

a bypass controller coupled to the data path circuit to control the first selection circuit based at least in part on an operating frequency of the data path circuit.

2. The apparatus of claim 1 , wherein the bypass controller comprises a path delay monitor having:
a replica circuit comprising one or more pipeline stages and at least one bypassed pipeline stage; and
a delayed replica circuit having at least one buffer and at least one pipeline stage.

3. The apparatus of claim 2 , further comprising a second selection circuit having a first input to receive an input to another pipeline stage and a second input to receive an output of the another pipeline stage.

4. The apparatus of claim 3 , wherein the bypass controller is to provide a first control signal to control the first selection circuit and a second control signal to control the second selection circuit.

5. The apparatus of claim 2 , wherein the path delay monitor further comprises a logic circuit to compare an output of the replica circuit and an output of the delayed replica circuit, wherein the bypass controller is to control the first selection circuit based on a comparison result from the logic circuit.

6. The apparatus of claim 1 , further comprising:
at least one core to execute instructions;
a cache memory coupled to the at least one core; and
an interconnect to couple a first circuit to a second circuit, wherein the interconnect comprises the data path circuit.

7. The apparatus of claim 6 , wherein the data path circuit comprises an RC dominated delay path and wherein the at least one core comprises one or more gate dominated delay paths.

8. The apparatus of claim 7 , wherein the bypass controller is to re-configure the RC dominated delay path in response to an update to the operating frequency, and wherein the one or more gate dominated delay paths are to be statically configured.

9. The apparatus of claim 6 , wherein the first circuit comprises the at least one core and the second circuit comprises the cache memory.

10. The apparatus of claim 2 , wherein the bypass controller is to enable the path delay monitor in response to an update to the operating frequency.

11. At least one computer readable storage medium having stored thereon instructions, which if performed by a machine cause the machine to perform a method comprising:
in response to an update to an operating frequency of at least a portion of a processor, enabling a path delay monitor associated with a data path circuit of the processor;
determining, in the path delay monitor, whether there is sufficient timing margin at the updated operating frequency to remove one or more pipeline stages of the data path circuit comprising identifying the sufficient timing margin when a first output of a replica circuit of the path delay monitor matches a second output of a delayed replica circuit of the path delay monitor; and
in response to determining that there is the sufficient timing margin, causing the one or more pipeline stages of the data path circuit to be bypassed.

12. The at least one computer readable storage medium of claim 11 , wherein the method further comprises in response to determining that there is insufficient timing margin, maintaining a current configuration of the data path circuit.

13. The at least one computer readable storage medium of claim 11 , wherein the method further comprises identifying an insufficient timing margin when the first output of the replica circuit does not match the second output of the delayed replica circuit.

14. The at least one computer readable storage medium of claim 11 , wherein the method further comprises draining the data path circuit prior to causing the one or more pipeline stages of the data path circuit to be bypassed.

15. The at least one computer readable storage medium of claim 11 , wherein the method further comprises stalling a producer circuit that provides a data input to the data path circuit prior to causing the one or more pipeline stages of the data path circuit to be bypassed.

16. A system comprising:
a processor comprising:
at least one core having a logic circuit to perform an operation on data, the logic circuit comprising a logic gate delay dominated path;
a cache memory coupled to the at least one core;
an interconnect to couple the at least one core to the cache memory, the interconnect having a RC delay dominated path comprising:
a plurality of pipeline stages to receive the data and output the data with a controllable amount of delay; and
a first selection circuit coupled between a first pipeline stage and a second pipeline stage, the first selection circuit having a first input to receive an input to the first pipeline stage and a second input to receive an output of the first pipeline stage and controllable to output one of the input to the first pipeline stage and the output of the first pipeline stage; and

a bypass controller coupled to the interconnect to control the first selection circuit based at least in part on an operating frequency of at least a portion of the processor; and

a system memory coupled to the processor.

17. The system of claim 16 , wherein the bypass controller comprises a path delay monitor having:
a replica circuit comprising one or more pipeline stages and at least one bypassed pipeline stage; and
a delayed replica circuit having at least one buffer and at least one pipeline stage.

18. The system of claim 17 , further comprising a second selection circuit having a first input to receive an input to another pipeline stage and a second input to receive an output of the another pipeline stage.

19. The system of claim 16 , wherein the interconnect comprises a data bus, and wherein the cache memory has a first latency to return data items when the interconnect is in operation at a first voltage and a second latency to return the data items when the interconnect is in operation at a second voltage.

Assignees

• Intel Corp

Inventors

• Carlson Ryan
• Dai Jianwei

Classifications

• Y02D10/00

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

• G06F1/3296

  by lowering the supply or operating voltage · CPC title

• G06F9/30079

  Pipeline control instructions, e.g. multicycle NOP · CPC title

• G06F1/3206

  Monitoring of events, devices or parameters that trigger a change in power modality · CPC title

• G06F9/3875

  Pipelining a single stage, e.g. superpipelining · CPC title