# US 2006/0155964 A1 · Method and apparatus for enable/disable control of SIMD processor slices

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2006/0155964 A1
- Document ID: US20060155964A1
- Status: Published application
- Kind: A1
- Application number: 11/034,693
- Filed: 2005-01-13
- Published: 2006-07-13
- Earliest claimed: 2005-01-13
- Assignees: Sony Computer Entertainment Inc.
- Inventors: Yonetaro Totsuka
- Web page: https://patentgrep.com/patent/US20060155964A1

## CPC

- G06F 9/30101: {Special purpose registers}
- G06F 1/3287: by switching off individual functional units in the computer system
- G06F 9/30072: {to perform conditional operations, e.g. using predicates or guards}
- G06F 9/30145: {Instruction analysis, e.g. decoding, instruction word fields}
- G06F 9/3838: {Dependency mechanisms, e.g. register scoreboarding}
- G06F 9/3858: {Result writeback, i.e. updating the architectural state or memory}
- G06F 9/3887: {controlled by a single instruction for multiple data lanes [SIMD]}
- Y02D 10/00: Energy efficient computing, e.g. low power processors, power management or thermal management
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- Y02D: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE

## Abstract

Methods and apparatus provide for disabling at least some data path processing circuits of a SIMD processing pipeline, in which the processing circuits are organized into a matrix of slices and stages, in response to one or more enable flags during a given cycle.

## Claims

### Claim 1

An apparatus, comprising a SIMD processing pipeline having a plurality of data path processing circuits organized into a matrix of slices and stages, wherein at least some of the processing circuits are operable to at least partially disable operation in response to one or more enable flags during a given cycle.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the processing circuits are operable to independently disable operation in response to respective ones of the enable flags.

### Claim 3 (depends on 1)

The apparatus of claim 1, further comprising: a clock interrupt circuit responsive to states of the enable flags by interrupting clock signals to the respective processing circuits.

### Claim 4 (depends on 1)

The apparatus of claim 1, further comprising: one or more control circuits operable to at least one of: (i) disable respective data results from the respective processing circuits from being written to one or more destination registers as a function of the enable flags; and (ii) enable the forwarding of results from the processing circuits to the one or more source operands to the pipeline.

### Claim 5 (depends on 4)

The apparatus of claim 4, wherein at least one of:
the one or more control circuits are operable to disable the writing of unused slice words in the results from the processing circuits to the destination register;
the one or more control circuits are operable to enable the forwarding of used slice words in the results from the processing circuits to the one or more source operands; and
the one or more control circuits are operable to perform the disable and/or enable function on a cycle-by-cycle basis.

### Claim 6 (depends on 4)

The apparatus of claim 4, further comprising:
a dependency check circuit operable to determine whether any operands of instructions yet to enter the slices of the pipeline are dependent on any of the data results of the processing circuits,
wherein the one or more control circuits are operable to disable or enable the forwarding function of the respective processing circuits based on the determinations of the dependency check circuit and the enable flags.

### Claim 7 (depends on 1)

The apparatus of claim 1, further comprising: a storage circuit operable to store respective sets of enable flags, each set having a respective enable flag for each processing circuit in a respective one of the stages.

### Claim 8 (depends on 7)

The apparatus of claim 7, wherein the storage circuit is operable to transfer the sets of enable flags from one stage for use with the processing circuits of a next stage each cycle.

### Claim 9 (depends on 7)

The apparatus of claim 7, wherein each set of enable flags is associated with a respective SIMD instruction.

### Claim 10 (depends on 7)

The apparatus of claim 7, wherein the respective sets of enable flags are at least one of:
contained within each SIMD instruction and are extracted and transferred to the storage circuit;
contained within respective SIMD control instructions, each control instruction being associated with a respective one of the SIMD instructions, and the enable flags being produced from the SIMD control instructions and transferred to the storage circuit;
produced in response to at least one of an absence of enable flags within the SIMD instructions and an absence of SIMD control instructions by selecting a predetermined set of bits; and
produced in response to information decoded from at least one of the SIMD control instructions and the SIMD instructions and used to select a predetermined set of enable flags.

### Claim 11 (depends on 10)

The apparatus of claim 10, wherein the SIMD control instructions are processed at least one of before, during, and after the SIMD instructions.

### Claim 12 (depends on 10)

The apparatus of claim 10, further comprising a storage circuit operable to store a plurality of predetermined sets of enable flags, the information decoded from at least one of the SIMD control instructions and the SIMD instructions being used to select one of the predetermined sets of enable flags.

### Claim 13

A method, comprising: disabling at least some data path processing circuits of a SIMD processing pipeline, in which the processing circuits are organized into a matrix of slices and stages, in response to one or more enable flags during a given cycle.

### Claim 14 (depends on 13)

The method of claim 13, wherein the step of disabling includes independently disabling the processing circuits in response to respective ones of the enable flags.

### Claim 15 (depends on 13)

The method of claim 13, wherein the step of disabling includes interrupting clock signals to the respective processing circuits in response to states of the enable flags.

### Claim 16 (depends on 13)

The method of claim 13, wherein at least one of:
the step of disabling includes preventing respective data results from the respective processing circuits from being written to one or more destination registers as a function of the enable flags; and
the method further comprises enabling the forwarding of results from the processing circuits to the one or more source operands to the pipeline.

### Claim 17 (depends on 16)

The method of claim 16, wherein at least one of:
the step of disabling includes disabling the writing of unused slice words in the results from the processing circuits to the destination register;
the step of forwarding includes enabling the forwarding of used slice words in the results from the processing circuits to the one or more source operands; and
the step of disabling and/or forwarding is performed on a cycle-by-cycle basis.

### Claim 18 (depends on 16)

The method of claim 16, wherein the steps of disabling and/or enabling are based on determinations of a dependency check circuit and the enable flags, the dependency check circuit being operable to determine whether any operands of instructions yet to enter the slices of the pipeline are dependent on any of the data results of the processing circuits.

### Claim 19 (depends on 13)

The method of claim 13, further comprising: storing respective sets of enable flags, each set having a respective enable flag for each processing circuit in a respective one of the stages.

### Claim 20 (depends on 19)

The method of claim 19, further comprising transferring the sets of enable flags from one stage for use with the processing circuits of a next stage each cycle.

### Claim 21 (depends on 20)

The method of claim 20, wherein each set of enable flags is associated with a respective SIMD instruction.

### Claim 22 (depends on 20)

The method of claim 20, further comprising at least one of:
providing the respective sets of enable flags within respective SIMD instructions and extracting the enable flags for storage;
providing the respective sets of enable flags within respective SIMD control instructions, associating each control instruction with a respective one of the SIMD instructions, and producing the enable flags from the SIMD control instructions for storage;
decoding the respective SIMD instructions and selecting a predetermined set of bits to produce the respective sets of enable flags when at least one of a given SIMD instruction does not include enable flags and there are no SIMD control instructions; and
decoding information from at least one of the SIMD control instructions and the SIMD instructions and using the information to select a predetermined set of enable flags.

### Claim 23

A storage medium containing an executable program, the executable program being operable to cause a processing system to execute actions including: disabling at least some data path processing circuits of a SIMD processing pipeline of the processing system, in which the processing circuits are organized into a matrix of slices and stages, in response to one or more enable flags during a given cycle.

### Claim 24 (depends on 23)

The storage medium of claim 23, wherein the step of disabling includes independently disabling the processing circuits in response to respective ones of the enable flags.

### Claim 25 (depends on 23)

The storage medium of claim 23, wherein the step of disabling includes interrupting clock signals to the respective processing circuits in response to states of the enable flags.

### Claim 26 (depends on 23)

The storage medium of claim 23, wherein at least one of:
the step of disabling includes preventing respective data results from the respective processing circuits from being written to one or more destination registers as a function of the enable flags; and
the actions further comprise enabling the forwarding of results from the processing circuits to the one or more source operands to the pipeline.

### Claim 27 (depends on 26)

The storage medium of claim 26, wherein at least one of:
the step of disabling includes disabling the writing of unused slice words in the results from the processing circuits to the destination register;
the step of forwarding includes enabling the forwarding of used slice words in the results from the processing circuits to the one or more source operands; and
the step of disabling and/or forwarding is performed on a cycle-by-cycle basis

### Claim 28

An apparatus, comprising:
a SIMD processing pipeline having a plurality of data path processing circuits organized into a matrix of slices and stages; and
a forwarding circuit operable to enable or disable the forwarding of results from the processing circuits to the one or more source operands to the pipeline in response to one or more enable flags during a given cycle.

### Claim 29 (depends on 28)

The apparatus of claim 28, wherein at least one of:
the forwarding circuit is operable to enable the forwarding of used slice words in the results from the processing circuits to the one or more source operands;
the forwarding circuit is operable to disable the forwarding of unused slice words in the results from the processing circuits to the one or more source operands;
the forwarding circuit is operable to perform the disable and/or enable function on a slice-by-slice basis;
the forwarding circuit is operable to perform the disable and/or enable function on a cycle-by-cycle basis.

### Claim 30 (depends on 28)

The apparatus of claim 28, further comprising:
a dependency check circuit operable to determine whether any operands of instructions yet to enter the slices of the pipeline are dependent on any of the data results of the processing circuits,
wherein the forwarding circuit is operable to disable or enable the forwarding function of the respective processing circuits based on the determinations of the dependency check circuit and the enable flags.

### Claim 31 (depends on 28)

The apparatus of claim 28, wherein each set of enable flags is associated with a respective SIMD instruction.

### Claim 32 (depends on 31)

The apparatus of claim 31, wherein the respective sets of enable flags are at least one of:
contained within each SIMD instruction and are extracted and transferred to the storage circuit;
contained within respective SIMD control instructions, each control instruction being associated with a respective one of the SIMD instructions, and the enable flags being produced from the SIMD control instructions and transferred to the storage circuit;
produced in response to at least one of an absence of enable flags within the SIMD instructions and an absence of SIMD control instructions by selecting a predetermined set of bits; and
produced in response to information decoded from at least one of the SIMD control instructions and the SIMD instructions and used to select a predetermined set of enable flags.
