Re-triggering wake-up to handle time skew between scalar and vector sides
US-2024184588-A1 · Jun 6, 2024 · US
US9348406B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-9348406-B2 |
| Application number | US-201213457011-A |
| Country | US |
| Kind code | B2 |
| Filing date | Apr 26, 2012 |
| Priority date | Jan 25, 2008 |
| Publication date | May 24, 2016 |
| Grant date | May 24, 2016 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
The present invention discloses a method of processing instructions in a pipeline-based central processing unit, wherein the pipeline is partitioned into base pipeline stages and enhanced pipeline stages according to functions, the base pipeline stages being activated all the while, and the enhanced pipeline stages being activated or shutdown according to requirements for performance of a workload. The present invention further discloses a method of processing instructions in a pipeline-based central processing unit, wherein the pipeline is partitioned into base pipeline stages and enhanced pipeline stages according to functions, each pipeline stage being partitioned into a base module and at least one enhanced module, the base module being activated all the while, and the enhanced module being activated or shutdown according to requirements for performance of a workload.
Opening claim text (preview).
The invention claimed is: 1. A method for a pipeline-based processor, wherein the pipeline-based processor includes at least one pipeline, the at least one pipeline including a particular pipeline, the method comprising: partitioning the particular pipeline into a plurality of pipeline stages based on one or more functions, the plurality of pipeline stages including one or more base pipeline stages and a plurality of enhanced pipeline stages; partitioning a particular pipeline stage of the plurality of pipeline stages into at least one base module and a plurality of enhanced modules, each base module of the at least one base module being a partitioned portion of the particular pipeline stage of the particular pipeline of the pipeline-based processor, each enhanced module of the plurality of enhanced modules being a partitioned portion of the particular pipeline stage of the particular pipeline of the pipeline-based processor, each enhanced module of the particular pipeline stage having a particular priority level of at least two different priority levels, at least one of the plurality of enhanced modules of the particular pipeline stage having a different priority level than at least one other of the plurality of enhanced modules, the at least two different priority levels including at least a highest priority level, the highest priority level having a relative priority higher than any other priority level of the at least two different priority levels, each of the at least one base module of the particular pipeline stage configured to be activated while the particular pipeline stage is activated; configuring each of the plurality of enhanced modules to be activated or shut down based on the particular priority level of a corresponding particular enhanced module and requirements for performance of a workload, wherein at least one of the plurality of enhanced modules of the particular pipeline stage is configured to be activated while at least another of the plurality of enhanced modules of the particular pipeline stage is shut down; determining each enhanced module of the particular pipeline stage to have one particular priority level of the at least two different priority levels according to a performance-to-power consumption ratio of a particular enhanced module, wherein a relatively high performance-to-power consumption ratio corresponds to a relatively high priority level; and configuring two or more of the plurality of enhanced modules of the particular pipeline stage to be activated in a high-to-tow order of the at least two different priority levels when the workload increases. 2. The method of claim 1 , further comprising: configuring a lookup table to store priority level information of the plurality of enhanced modules of the particular pipeline stage. 3. The method of claim 1 , wherein a particular enhanced module of the particular pipeline stage having the highest priority level has a higher performance-to-power consumption ratio than any other enhanced module having another priority level. 4. The method of claim 1 , wherein a base module of a particular base pipeline stage executes a smallest function of the particular base pipeline stage. 5. The method of claim 1 , wherein an enhanced module of a particular base pipeline stage enhances one or more functions of the particular base pipeline stage. 6. The method of claim 1 , wherein a base module of a particular enhanced pipeline stage enables data to pass through the particular enhanced pipeline stage without being changed. 7. The method of claim 1 , wherein each enhanced pipeline stage has an enhanced stage priority level of at least two different enhanced stage priority levels, the at least two different enhanced stage priority levels including at least a highest enhanced stage priority level, the highest enhanced stage priority level having a relative enhanced stage priority higher than any other enhanced stage priority level of the at least two different enhanced stage priority levels, the method further comprising: configuring each enhanced pipeline stage of the plurality of enhanced pipeline stages to be activated or shut down based on the enhanced stage priority level of each enhanced pipeline stage and requirements for performance of the workload. 8. The method of claim 1 , further comprising: configuring each enhanced pipeline stage of the plurality of enhanced pipeline stages to be activated or shut down. 9. The method of claim 1 , wherein partitioning a particular pipeline stage of the plurality of pipeline stages into at least one base module and a plurality of enhanced modules further comprises: partitioning a base pipeline stage into at least one base module and plurality of enhanced modules. 10. The method of claim 1 , wherein partitioning a particular pipeline stage of the plurality of pipeline stages into at least one base module and a plurality of enhanced modules further comprises: partitioning an enhanced pipeline stage into at least one base module and a plurality of enhanced modules. 11. The method of claim 1 , wherein each enhanced pipeline stage is configured to be either a shutdown enhanced pipeline stage or an activated enhanced pipeline stage, all functions of each shutdown enhanced pipeline stage being not activated, at least one function of each activated enhanced pipeline stage being activated, wherein at least one of the plurality of enhanced pipeline stages is configured to be activated while at least another of the plurality of enhanced pipeline stages is shut down. 12. A pipeline-based processor, comprising: at least one pipeline, the at least one pipeline at least including a particular pipeline, wherein the particular pipeline is partitioned into a plurality of pipeline stages based on one or more functions, the plurality of pipeline stages including one or more base pipeline stages and a plurality of enhanced pipeline stages, a particular pipeline stage of the plurality of pipeline stages including at least one base module and a plurality of enhanced modules, each base module of the at least one base module being a partitioned portion of the particular pipeline stage of the particular pipeline of the pipeline-based processor, each enhanced module of the plurality of enhanced modules being a partitioned portion of the particular pipeline stage of the particular pipeline of the pipeline-based processor, each of the at least one base module of the particular pipeline stage being activated while the particular pipeline stage is activated, each enhanced module of the particular pipeline stage having a particular priority level of at least two different priority levels, at least one of the plurality of enhanced modules of the particular pipeline stage having a different priority level than at least one other of the plurality of enhanced modules, the at least two different priority levels including at least a highest priority level, the highest priority level having a relative priority higher than any other priority level of the at least two different priority levels, wherein each priority level is determined according to a performance-to-power consumption ratio of a particular enhanced module of the particular pipeline stage, wherein a relatively high performance-to-power consumption ratio corresponds to a relatively high priority level; and a controller circuit, the controller circuit being configured to activate or shut down each enhanced module of the plurality of enhanced modules of the particular pipeline stage based on the particular priority level of a corresponding particular enhanced module of the particular pipeline stage and requirements for performance of a workload, wherein
using instruction pipelines · CPC title
by switching off individual functional units in the computer system · CPC title
using a plurality of independent parallel functional units · CPC title
Monitoring · CPC title
based on priority control (G06F13/1605 takes precedence) · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.