GB2486485A - Completing execution of one phase of a computer program before scheduling any instructions for the next phase - Google Patents

Completing execution of one phase of a computer program before scheduling any instructions for the next phase Download PDF

Info

Publication number
GB2486485A
GB2486485A GB1021414.6A GB201021414A GB2486485A GB 2486485 A GB2486485 A GB 2486485A GB 201021414 A GB201021414 A GB 201021414A GB 2486485 A GB2486485 A GB 2486485A
Authority
GB
United Kingdom
Prior art keywords
program
phase
thread
scheduler
microprocessor
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
GB1021414.6A
Other versions
GB2486485B (en
GB201021414D0 (en
Inventor
Yoong Chert Foo
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Imagination Technologies Ltd
Original Assignee
Imagination Technologies Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Imagination Technologies Ltd filed Critical Imagination Technologies Ltd
Priority to GB1021414.6A priority Critical patent/GB2486485B/en
Publication of GB201021414D0 publication Critical patent/GB201021414D0/en
Priority to US13/068,791 priority patent/US9304812B2/en
Priority to PCT/GB2011/052463 priority patent/WO2012080720A1/en
Priority to EP11808288.2A priority patent/EP2652597B1/en
Priority to CN201180065275.8A priority patent/CN103314357B/en
Publication of GB2486485A publication Critical patent/GB2486485A/en
Application granted granted Critical
Publication of GB2486485B publication Critical patent/GB2486485B/en
Priority to US15/056,866 priority patent/US10585700B2/en
Priority to US16/805,727 priority patent/US11947999B2/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/48—Program initiating; Program switching, e.g. by interrupt
    • G06F9/4806—Task transfer initiation or dispatching
    • G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
    • G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/76—Architectures of general purpose stored program computers
    • G06F15/78—Architectures of general purpose stored program computers comprising a single central processing unit
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F15/00—Digital computers in general; Data processing equipment in general
    • G06F15/76—Architectures of general purpose stored program computers
    • G06F15/80—Architectures of general purpose stored program computers comprising an array of processing units with common control, e.g. single instruction multiple data processors
    • G06F15/8007—Architectures of general purpose stored program computers comprising an array of processing units with common control, e.g. single instruction multiple data processors single instruction multiple data [SIMD] multiprocessors
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/30003—Arrangements for executing specific machine instructions
    • G06F9/3005—Arrangements for executing specific machine instructions to perform operations for flow control
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/30003—Arrangements for executing specific machine instructions
    • G06F9/30076—Arrangements for executing specific machine instructions to perform miscellaneous control operations, e.g. NOP
    • G06F9/3009—Thread control instructions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
    • G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
    • G06F9/3836—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution
    • G06F9/3851—Instruction issuing, e.g. dynamic instruction scheduling or out of order instruction execution from multiple instruction streams, e.g. multistreaming
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
    • G06F9/3867—Concurrent instruction execution, e.g. pipeline, look ahead using instruction pipelines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
    • G06F9/3885—Concurrent instruction execution, e.g. pipeline, look ahead using a plurality of independent parallel functional units
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
    • G06F9/38—Concurrent instruction execution, e.g. pipeline, look ahead
    • G06F9/3885—Concurrent instruction execution, e.g. pipeline, look ahead using a plurality of independent parallel functional units
    • G06F9/3887—Concurrent instruction execution, e.g. pipeline, look ahead using a plurality of independent parallel functional units controlled by a single instruction for multiple data lanes [SIMD]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/52—Program synchronisation; Mutual exclusion, e.g. by means of semaphores
    • G06F9/522—Barrier synchronisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T1/00—General purpose image data processing
    • G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • 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
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

A program for a microprocessor, such as a single instruction, multiple data (SIMD) graphics processing unit, is divided into separate phases. The processor schedules the instructions in one phase and waits for all the instructions in that phase to complete before scheduling the instructions in the next phase. The instructions may be scheduled on multiple threads, each thread processing one or more data instances. The processor may initialise a counter to the number of threads being executed in the program phase and decrement the counter each time a thread completes. In this case, the processor schedules the next phase of the program when the counter reaches zero. The number of threads used for a given phase may be configured dynamically based on the results of the preceding phase.

Description

I
Method and Apparatus for ScheduIiq the Issue of Instructions in a Microprocessor usinQMultipe Phases of Execution
Field of the invention
The present invention relates to the field of microprocessors and is particularly advantageous for single instruction multiple data (SIMD) processors.
Background to the inventiofl
With some types of data, such as graphics data, large blocks of data often need to undergo the same processing operations. One example is when changing the brightness of an image. SIMD processing is advantageous for this process. An SIMD processor is a microprocessor with execution units, caches and memories as with any other processor, but additionally incorporates the concept of parallel execution of multiple data streams. Instead of individually fetching data instances, a single fetch instruction can be used to fetch a block of data. A single instruction can then be used to operate on the fetched data block, such as an "add" operation. SIMD processing therefore reduces the load on the processor as compared with traditional single instruction single data (SISD) processing.
However, SIMD processing can only be used in limited circumstances. For example, even in situations where large blocks of data are initially to be processed in the same way, programs often contain conditional instructions or branch instructions that result in some data in a block being operated on by one branch of instructions and the remaining data by another branch of instructions.
It is very often not possible to predict in advance how many data instances will need to be processed by one branch and how many by another.
Typically in an SIMD processor, a scheduler is used to schedule the execution of a program, allocating the resources required by the program at the outset.
One solution for programs in which different parts of a data set are processed by different instruction during a portion of the program has been to execute each branch on all of the data and then discard the unwanted results. Clearly this is an inefficient use of processor resources and time.
Accordingly, it would be desirable to be able better allocate resources during s the execution of programs in an SIMD processor, both to reduce power consumption and to optimise resource usage.
Summariof the invention In a first aspect, the invention provides a microprocessor configured to execute programs divided into discrete phases, comprising: a scheduler for scheduling program instructions to be executed on the processor; and a plurality of resources for executing programming instructions issued is by the scheduler; wherein the scheduler is configured to schedule each phase of the program only after receiving an indication that execution of the preceding phase of the program has been completed.
Preferably, the processor is a multithreaded processor. Preferably, the processor is an SIMD processor. Preferably, the scheduler is configured to recalculate an SIMD ratio for each phase of the program. The SIMD ratio is the the number of data instances processed by a single instruction.
The processor includes a feedback loop to the scheduler through which the completion of each phase of the program is communicated to the scheduler.
Preferably the scheduler maintains a record for each phase of the program that is scheduled, during execution of that phase of the program.
Preferably, the processor further includes a thread finished counter, wherein when each thread finishes a phase of the program an indication is sent to the thread finished counter, and the processor is configured such that the scheduler schedules a next phase of the program only when the thread finished counter indicates that all threads for a preceding phase of the program have finished. The thread finished counter may be part of the scheduler or provided as a separate component. The thread finished counter may comprise a memory storing a number of threads for each phase of a program and the thread finished counter decremented each time a thread finishes a phase, and configured such that when the thread finished counter reaches zero, the s scheduler is instructed to schedule the next phase of the program.
The thread finished counter may store a record of a number of data instances for each thread of a phase of a program.
Preferably, the scheduler is configured to dynamically allocate the number of threads for each phase of the program based on the results of a preceding phase.
In a second aspect, the invention provides a method for scheduling programs in a microprocessor, the microprocessor comprising a scheduler for scheduling programs of programming instructions, the programs being divided into discrete phases, the method comprising the steps of: scheduling a first phase of a program to be executed on the processor; executing the first phase of the program scheduled by the scheduler; when execution of the first phase of the program is complete, providing an indication to the scheduler that execution of the first phase of the program is complete; scheduling a second phase of the program after the scheduler has received the indication that execution of the first phase of the program is complete.
Preferably, the method further comprises maintaining a record for each phase of the program that is scheduled, during execution of that phase of the program.
Preferably, the method further comprises updating the record when each thread finishes a phase of the program, and scheduling a next phase of the program only when the record indicates that all threads for a preceding phase of the program have finished.
Preferably, the method further comprises storing a record of a number of data instances for each thread of a phase of a program.
Preferably, the method further comprises dynamically allocating the number of threads for each phase of the program based on the results of a preceding phase.
In a third aspect, the invention provides a computer program product, comprising computer executable code in the form a program executable on an io SIMD processor, wherein the program is divided into a plurality of phases by phase instructions, the phase instructions being provided at points in the program where branches may occur and allowing a scheduler to schedule each phase of the program separately. At these points the SIMD ratio may need to change. An example of a point at which the SIMD ratio may need to change is following a branch instruction or conditional instruction. Another example is a sub-routine for alpha blending or anti-aliasing in a graphics processing program.
Brief Description of the Drawings
Embodiments of the invention will now be described in detail, by way of example only, with reference to the accompanying drawings, in which: Figure 1 is a schematic illustration of a processor in accordance with the present invention; Figure 2 is a schematic illustration of a program scheduling scheme in accordance with the present invention; Figure 3 is a flow diagram illustrating the steps carried out by the scheduler of Figure 2; and Figure 4 is a flow diagram illustrating the process carried out by the scheduler for scheduling new threads.
Detailed Description
Figure 1 is a schematic illustration of a multi-threaded SIMD processor 10 in accordance with an embodiment of the present invention. The processor is configured to schedule and execute individual phases of each program, separately to one another, rather than scheduling a complete program all at once and then executing it.
As with conventional processors, following scheduling by the scheduler 100, instructions are fetched from memory by the instruction fetch units 110.
Fetched instructions are decoded by the instruction decode unit 120 and fed to the issue logic 130. The issue logic feeds the instructions and fetched data to the arithmetic logic units (ALUs) 140 and the results are stored in registers or memory 150.
The ALUs 140 are also configured to feedback thread finished instructions to the scheduler 100. As will be explained, thread finished instructions are provided at the end of each phase of a program.
New programs are provided to the scheduler 100 divided into multiple phases.
The way in which the programs are divided is determined by the programmer but typically a new phase will begin at a point in the program when the resource requirements of the program may change.
For example, a program may include a conditional instruction such that if a data value is less than a threshold it is processed by one branch of instructions otherwise it is processed by another branch of instructions. It may not be possible to determine whether all data instances being processed by the program will be processed by a single branch or whether the data will be split between the two branches and if so, how many data instances will go down which branch. So at the point of the conditional instruction the number of threads required and the number of instruction fetches required may change in a way that cannot be reliably predicted. By splitting a program at such points, resource scheduling can be managed dynamically in response to the data results.
Another example of an advantageous point for a new program phase is a routine for blending background colour with a transparent foreground object in a graphics program. A foreground object may have a uniform colour and so the data instances for the entire object can be processed in the same way for a colour calculation. However, the background over which the object sits may have variable colour and so different portions of the object must be processed differently in a blending routine. The blending routine may therefore be advantageously placed in a separate phase of the program to the colour routine.
A further example of an advantageous point for a new program phase is an anti-aliasing routine. Anti-aliasing is used when representing a high resolution image on a lower resolution screen. To remove undesirable artefacts, such as jagged edges, an anti-aliasing routine is used to smooth out the image. But anti-aliasing is only required at the edges of graphics objects. If a block of uniform colour is in the image, it is not necessary to process the bulk of it at high resolution and perform anti-aliasing. Only at the boundaries with other objects is it necessary. So a program might include a branch instruction so that edge portions of image data are processed using the anti-aliasing routine while larger blocks of data are processed in a much less computationally expensive way.
The programs are divided into phases by the inclusion of "phase completed" instructions at the end of each phase of the program. The phase completed instructions when executed feed back to the scheduler 100 an indication that a phase as been completed. The phase completed instructions include a program identifier.
The program instructions also include an indication of the resource requirements of the program, but instead of being provided at the start of the program to cover the entire program, they are provided at the beginning of each phase of the program. The scheduler 100 can then assess if there are sufficient resources available to execute the phase. Because the resources required may depend on the outcome of a preceding phase, the indication of the resources required may be in the form an algorithm to calculate the resources required.
This dynamic calculation of resource requirements removes the need to always provision for the worst case scenario and leads to more efficient resource allocation.
s The scheduler 100 includes a multi-phased execution control unit or thread finished counter 160 that stores a record for each program scheduled, and the number of threads scheduled for the current phase. At the start of a new program the scheduler creates a new record in the thread finished counter 160, and updates it every time a new phase of that program is scheduled. As each io thread finishes a phase the thread finished counter receives an indication and adjusts the record for the number of threads executing for that program. The count of threads for a program is simply decremented each time a thread finishes until the thread count reaches zero. When the thread count for a program reaches zero, a new phase of that program can be scheduled.
The thread finished counter may be provided as part of the scheduler or as a separate component.
Figure 2 is a schematic illustration of scheduling and execution process in accordance with the present invention showing the content of a program queue and a thread finished counter. A stream of programs is provided in a program queue as an input to a scheduler. The programs 200 are each divided into phases as described above, and each has a predetermined number of data instances, indicated as the value I in the program queue. The number of instances per thread is indicated by the value R. The scheduler 100 schedules individual phases of a program and waits for an indication that that phase is complete before scheduling the next phase. In the example shown in Figure 2 there are three programs in the program queue not yet scheduled and two programs, program A and program B currently being executed, but with further phases remaining to be scheduled. The number of instances per thread, R, is known for the initial phase of each program but may be altered for subsequent phases of the program depending on the outcome of preceding phases. So the value of R is not known for the phases of programs A and B yet to be scheduled.
The phase of program A being executed has 17 threads, indicated in the thread finished counter 160 by the value T. Each thread contains a single data instance, so R=1. The phase of program B being executed has 2 threads each with 5 instances per thread. Figure 2 shows multiple threads scheduled by the scheduler entering the execution pipeline 210. The execution pipeline comprises the instruction fetch, issue logic and ALU's shown in Figure 1.
As already described with reference to Figure 1, the thread finished counter stores a record for each program scheduled, and the number of threads scheduled for the current phase. As each thread finishes a phase the thread to finished counter 160 receives an indication 220 from the end of the execution pipeline 210 and adjusts the record for the number of threads executing for that program. The count of threads T for a program is simply decremented each time a thread finishes until the thread count reaches zero. When the thread count for a program reaches zero, a new phase of that program can be is scheduled.
Figure 3 is a flow diagram showing the process steps carried out by the scheduler in deciding how to schedule new threads. In step 300 the scheduler first checks if there are threads available for running to program from the program queue that are available to run. If there no threads available, the scheduler continues to check until there are threads available.
If there are threads available to run the first program in the queue is evaluated in step 310. In step 320, the scheduler checks if the first program already has a phase running and the scheduler is waiting for feedback to indicate that the phase is completed. If so then in step 325, the next program in the queue is evaluated in the same manner. If the scheduler is not awaiting feedback from a phase of the first program, then in step 330 the scheduler assesses whether there are any further phases to execute. If there are no further phases the program is removed from the queue, as shown in step 340, and its record removed from thread finished counter.
If there are more phases left to run the scheduler determines whether the program is still running in step 350. If the program is still running, the scheduler moves to step 320 and next program in the queue is evaluated. If the program is not still running, the scheduler creates new threads for the next phase of the program, as indicated by step 360. Following the creation of new threads, the scheduler returns to the step 300 to check if there is room for any further threads.
Figure 4 shows in detail the steps taken by the scheduler to create new threads in step 360. In a first step, step 400, the scheduler checks that there is space for a new thread. If not, the scheduler continues to check until space becomes available. When space for a new thread is available, the scheduler creates a io new thread, shown as step 410. The number of data instances in the new thread is the lesser of the maximum number of instances allowed per thread (which is limited by the hardware) and the number of data instances left.
Once the new thread has been created, a record for the thread is created or updated in the multi phased execution control unit, shown as step 420.
In step 430 the scheduler assesses whether there are further data instances to be run in the phase. If so, the scheduler attempts to create a further thread or threads to accommodate them. If not, the thread creation process ends in step 440 and the scheduler returns to step 300.
The present invention allows the SIMD ratio of programs to be dynamically altered during program execution. By splitting programs into multiple phases and providing a scheduler that is able to determine whether execution of a phase has been completed, each phase can be separately scheduled and the results of preceding phases can be used to inform the scheduling of subsequent phases.

Claims (16)

  1. Claims 1. A microprocessor configured to execute programs divided into discrete phases, comprising: a scheduler for scheduling program instructions to be executed on the processor; a plurality of resources for executing programming instructions issued by the scheduler; wherein the scheduler is configured to schedule each phase of the io program only after receiving an indication that execution of the preceding phase of the program has been completed.
  2. 2. A microprocessor according to claim 1, wherein the microprocessor is a single instruction multiple data (SIMD) processor.
  3. 3. A microprocessor according to claim I or 2, wherein the scheduler maintains a record for each program that is scheduled, during execution of that program.
  4. 4. A microprocessor according to any preceding claim, wherein the microprocessor is a multithreaded processor and further includes a thread finished counter, wherein when each thread finishes a phase of the program an indication is sent to the thread finished counter, and wherein the processor is configured such that the scheduler schedules a next phase of the program only when the thread finish counter indicates that all threads for a preceding phase of the program have finished.
  5. 5. A microprocessor according to claim 4, wherein the thread finished counter is part of the scheduler.
  6. 6. A microprocessor according to claim 4 or 5, wherein the thread finished counter comprises a memory storing a number of threads for each phase of a program and the microprocessor is configured such that the thread finished counter is decremented each time a thread finishes a phase, and when the thread counter reaches zero, the scheduler is instructed to schedule the next phase of the program.
  7. 7. A microprocessor according to claim 4, 5 or 6, wherein the thread finished counter stores a record of a number of data instances for each thread of a phase of a program.
  8. 8. A microprocessor according to any preceding claim, wherein the microprocessor is configured such that the number of threads for each phase of the program is configured dynamically based on the results of a preceding io phase.
  9. 9. A method for scheduling programs in a microprocessor, the microprocessor comprising a scheduler for scheduling programs of programming instructions, the programs divided into discrete phases, the is method comprising the steps of: scheduling a first phase of a program to be executed on the processor; executing the first phase of the program scheduled by the scheduler; when execution of the first phase of the program is complete, providing an indication to the scheduler that execution of the first phase of the program is complete scheduling a second phase of the program after the scheduler has received the indication that execution of the first phase of the program is complete.
  10. 10. A method according to claim 9, further comprising maintaining a record for each phase of the program that is scheduled, during execution of that phase of the program.
  11. 11. A method according to claim 10, further comprising updating the record when each thread finishes a phase of the program, and scheduling a next phase of the program only when the record indicates that all threads for a preceding phase of the program have finished.
  12. 12. A method according to claim 9, 10 or 11, further comprising storing a 3s record of a number of data instances for each thread of a phase of a program.
  13. 13. A method according to any one of claims 9 to 12! further comprising dynamically allocating the number of threads for each phase of the program based on the results of a preceding phase.
  14. 14. A computer program product, comprising computer executable code in the form a computer program executable on an SIMD processor, the program being divided into a plurality of phases by phase instructions, the phase instructions being provided at points in the program where branches may occur io and allowing a scheduler to schedule each phase of the program separately.
  15. 15. A microprocessor substantially as described herein with reference to the accompanying drawings.is 16. A method for scheduling programs in a microprocessor substantially as described herein with reference to the accompanying drawings.AMENDMENTS TO THE CLAIMS HAVE BEEN FILED AS FOLLOWS1. A multithreaded single instruction multiple data (SIMD) microprocessor configured to execute programs divided into discrete phases of programming instructions, and configured such that a plurality of data instances can be processed by a single thread of programming instructions, comprising: a scheduler for scheduling phases of program instructions to be executed on the processor; and a plurality of resources for executing programming instructions issued by the scheduler; wherein the scheduler is configured to schedule each phase of the program only after receiving an indication that execution of a preceding phase of the program has been completed, and is configured during scheduling of each phase of the program to determine a number of data instances to be allocated to each thread for that phase of the program.2. A microprocessor according to any preceding claim, wherein the scheduler is configured to determine the number of data instances per thread for each phase of the program based on a result of a preceding phase of the program.3. A microprocessor according to claim I or 2, wherein the scheduler maintains a record for each program that is scheduled, during execution of that program.4. A microprocessor according to any preceding claim, further including a thread finished counter, wherein when each thread finishes a phase of the program an indication is sent to the thread finished counter, and wherein the processor is configured such that the *ns *.* : 25 scheduler schedules a next phase of the program only when the thread finish counter * ** *** * indicates that all threads for a preceding phase of the program have finished. * *. * * *5. A microprocessor according to claim 4, wherein the thread finished counter is part of the scheduler. ** *. 30 * ***6. A microprocessor according to claim 4 or 5, wherein the thread finished counter comprises a memory storing a number of threads for each phase of a program and the microprocessor is configured such that the thread finished counter is decremented each time a thread finishes a phase, and when the thread counter reaches zero, the scheduler is instructed to schedule the next phase of the program.7. A microprocessor according to claim 4, 5 or 6, wherein the thread finished counter stores a record of a number of data instances for each thread of a phase of a program.8. A method for scheduling programs in a multithreaded, single instruction multiple data microprocessor, the microprocessor comprising a scheduler for scheduling programs of programming instructions, the programs divided into discrete phases of programming instructions, the method comprising the steps of: scheduling a first phase of a program to be executed on the processor, including the step of determining how many data instances to allocate to each of a plurality of threads in the first phase; executing the first phase of the program scheduled by the scheduler; when execution of the first phase of the program is complete, providing an indication to the scheduler that execution of the first phase of the program is complete; and scheduling a second phase of the program after the scheduler has received the indication that execution of the first phase of the program is complete, including the step of determining how many data instances to allocate to each of a plurality of threads in the second phase.9. A method according to any one of claims 9 to 12, wherein the step of determining how many data instances to allocate to each thread for the second phase of the program is based on a result of the first phase of the program.10. A method according to claim 8 or 9, further comprising maintaining a record for each phase of the program that is scheduled, during execution of that phase of the program.11. A method according to claim 10, further comprising updating the record when each thread finishes a phase of the program, and scheduling a next phase of the program only when the record indicates that all threads for a preceding phase of the program have finished.12. A method according to claim 8, 9, 10 or 11, further comprising storing a record of a number of data instances for each thread of a phase of a program.13. A method according to any one of claims 8 to 12, further comprising dynamically allocating the number of threads for the second phase of the program based on a result of the first phase of the program.14. A computer program product, comprising computer executable code in the form.a S computer program executable on an SIMD processor, the program being divided into a plurality of phases by phase instructions, the phase instructions being provided at points in the program where branches may occur and allowing a scheduler to schedule each phase of the program separately and to determine how many data instances to allocate to each thread for each phase of the program.15. A microprocessor substantially as described herein with reference to the accompanying drawings.
  16. 16. A method for scheduling programs in a microprocessor substantially as described herein with reference to the accompanying drawings.SOS * SS* 55555 * S o so * 5 5 555 S * S o 055 *5 * * S * 5*
GB1021414.6A 2010-12-16 2010-12-16 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution Active GB2486485B (en)

Priority Applications (7)

Application Number Priority Date Filing Date Title
GB1021414.6A GB2486485B (en) 2010-12-16 2010-12-16 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution
US13/068,791 US9304812B2 (en) 2010-12-16 2011-05-19 Multi-phased and multi-threaded program execution based on SIMD ratio
PCT/GB2011/052463 WO2012080720A1 (en) 2010-12-16 2011-12-13 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution
EP11808288.2A EP2652597B1 (en) 2010-12-16 2011-12-13 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution
CN201180065275.8A CN103314357B (en) 2010-12-16 2011-12-13 Perform for using multiple stage in the microprocessor and the method and apparatus dispatched is issued to instruction
US15/056,866 US10585700B2 (en) 2010-12-16 2016-02-29 Multi-phased and multi-threaded program execution based on SIMD ratio
US16/805,727 US11947999B2 (en) 2010-12-16 2020-02-29 Multi-phased and multi-threaded program execution based on SIMD ratio

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
GB1021414.6A GB2486485B (en) 2010-12-16 2010-12-16 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution

Publications (3)

Publication Number Publication Date
GB201021414D0 GB201021414D0 (en) 2011-01-26
GB2486485A true GB2486485A (en) 2012-06-20
GB2486485B GB2486485B (en) 2012-12-19

Family

ID=43567365

Family Applications (1)

Application Number Title Priority Date Filing Date
GB1021414.6A Active GB2486485B (en) 2010-12-16 2010-12-16 Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution

Country Status (5)

Country Link
US (3) US9304812B2 (en)
EP (1) EP2652597B1 (en)
CN (1) CN103314357B (en)
GB (1) GB2486485B (en)
WO (1) WO2012080720A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2573316A (en) * 2018-05-02 2019-11-06 Advanced Risc Mach Ltd Data processing systems

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2486485B (en) 2010-12-16 2012-12-19 Imagination Tech Ltd Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution
US9697005B2 (en) * 2013-12-04 2017-07-04 Analog Devices, Inc. Thread offset counter
WO2023062456A1 (en) * 2021-10-14 2023-04-20 Braingines SA Dynamic, low-latency, dependency-aware scheduling on simd-like devices for processing of recurring and non-recurring executions of time-series data

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2006083291A2 (en) * 2004-06-08 2006-08-10 University Of Rochester Dynamically managing the communication-parallelism trade-off in clustered processors
US20080098208A1 (en) * 2006-10-24 2008-04-24 Arm Limited Analyzing and transforming a computer program for executing on asymmetric multiprocessing systems
US7542043B1 (en) * 2005-05-23 2009-06-02 Nvidia Corporation Subdividing a shader program
DE102008020782A1 (en) * 2008-04-25 2009-11-12 Fujitsu Siemens Computers Gmbh Processor i.e. arithmetic and logic unit, for use in CPU of computer system for parallel execution of two threads, has fork instruction for producing and starting thread, stop instruction for stopping current thread

Family Cites Families (29)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2311882B (en) 1996-04-04 2000-08-09 Videologic Ltd A data processing management system
CN100373331C (en) 1996-08-27 2008-03-05 松下电器产业株式会社 Multithreaded processor for processing multiple instruction streams independently of each other by flexibly controlling throughput in each instruction stream
US6128712A (en) * 1997-01-31 2000-10-03 Macromedia, Inc. Method and apparatus for improving playback of interactive multimedia works
WO1999024903A1 (en) 1997-11-07 1999-05-20 Bops Incorporated METHODS AND APPARATUS FOR EFFICIENT SYNCHRONOUS MIMD OPERATIONS WITH iVLIW PE-to-PE COMMUNICATION
US7518993B1 (en) 1999-11-19 2009-04-14 The United States Of America As Represented By The Secretary Of The Navy Prioritizing resource utilization in multi-thread computing system
JP3641997B2 (en) * 2000-03-30 2005-04-27 日本電気株式会社 Program conversion apparatus and method, and recording medium
US6732253B1 (en) 2000-11-13 2004-05-04 Chipwrights Design, Inc. Loop handling for single instruction multiple datapath processor architectures
JP3810631B2 (en) * 2000-11-28 2006-08-16 富士通株式会社 Recording medium on which information processing program is recorded
US7337442B2 (en) 2002-12-03 2008-02-26 Microsoft Corporation Methods and systems for cooperative scheduling of hardware resource elements
US7239322B2 (en) 2003-09-29 2007-07-03 Ati Technologies Inc Multi-thread graphic processing system
US7472390B2 (en) 2003-10-01 2008-12-30 Intel Corporation Method and apparatus to enable execution of a thread in a multi-threaded computer system
US8174531B1 (en) 2003-10-29 2012-05-08 Nvidia Corporation Programmable graphics processor for multithreaded execution of programs
US6897871B1 (en) 2003-11-20 2005-05-24 Ati Technologies Inc. Graphics processing architecture employing a unified shader
US7418576B1 (en) * 2004-11-17 2008-08-26 Nvidia Corporation Prioritized issuing of operation dedicated execution unit tagged instructions from multiple different type threads performing different set of operations
US7836116B1 (en) 2006-06-15 2010-11-16 Nvidia Corporation Fast fourier transforms and related transforms using cooperative thread arrays
US20080072015A1 (en) 2006-09-18 2008-03-20 Julier Michael A Demand-based processing resource allocation
US8018457B2 (en) 2006-09-19 2011-09-13 Caustic Graphics, Inc. Ray tracing system architectures and methods
US7969434B2 (en) 2006-09-19 2011-06-28 Caustic Graphics, Inc. Method, apparatus, and computer readable medium for accelerating intersection testing in ray-tracing rendering
US7984269B2 (en) 2007-06-12 2011-07-19 Arm Limited Data processing apparatus and method for reducing issue circuitry responsibility by using a predetermined pipeline stage to schedule a next operation in a sequence of operations defined by a complex instruction
US7478856B1 (en) * 2007-07-06 2009-01-20 Philmar, Llc Kit for conversion of horizontally hinged door of recreational vehicle to a table
US8996846B2 (en) 2007-09-27 2015-03-31 Nvidia Corporation System, method and computer program product for performing a scan operation
US8065288B1 (en) 2007-11-09 2011-11-22 Nvidia Corporation System, method, and computer program product for testing a query against multiple sets of objects utilizing a single instruction multiple data (SIMD) processing architecture
US7999808B1 (en) 2007-12-13 2011-08-16 Nvidia Corporation Parallel processing system, method, and computer program product for executing node traversal or primitive intersection
JP5510322B2 (en) * 2008-07-04 2014-06-04 日本電気株式会社 Multiprocessor system, multithread processing method, and program
US20100064291A1 (en) 2008-09-05 2010-03-11 Nvidia Corporation System and Method for Reducing Execution Divergence in Parallel Processing Architectures
US9269182B2 (en) 2008-09-05 2016-02-23 Nvidia Corporation System and method for identifying entry points of a hierarchical structure
US8424015B2 (en) * 2010-09-30 2013-04-16 International Business Machines Corporation Transactional memory preemption mechanism
US8499305B2 (en) * 2010-10-15 2013-07-30 Via Technologies, Inc. Systems and methods for performing multi-program general purpose shader kickoff
GB2486485B (en) * 2010-12-16 2012-12-19 Imagination Tech Ltd Method and apparatus for scheduling the issue of instructions in a microprocessor using multiple phases of execution

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2006083291A2 (en) * 2004-06-08 2006-08-10 University Of Rochester Dynamically managing the communication-parallelism trade-off in clustered processors
US7542043B1 (en) * 2005-05-23 2009-06-02 Nvidia Corporation Subdividing a shader program
US20080098208A1 (en) * 2006-10-24 2008-04-24 Arm Limited Analyzing and transforming a computer program for executing on asymmetric multiprocessing systems
DE102008020782A1 (en) * 2008-04-25 2009-11-12 Fujitsu Siemens Computers Gmbh Processor i.e. arithmetic and logic unit, for use in CPU of computer system for parallel execution of two threads, has fork instruction for producing and starting thread, stop instruction for stopping current thread

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2573316A (en) * 2018-05-02 2019-11-06 Advanced Risc Mach Ltd Data processing systems
US10861125B2 (en) 2018-05-02 2020-12-08 Arm Limited Preparing and executing command streams in data processing systems
GB2573316B (en) * 2018-05-02 2021-01-27 Advanced Risc Mach Ltd Data processing systems

Also Published As

Publication number Publication date
GB2486485B (en) 2012-12-19
US10585700B2 (en) 2020-03-10
US20200201678A1 (en) 2020-06-25
EP2652597A1 (en) 2013-10-23
US20160179519A1 (en) 2016-06-23
WO2012080720A1 (en) 2012-06-21
CN103314357B (en) 2016-01-20
EP2652597B1 (en) 2018-03-07
US20120159120A1 (en) 2012-06-21
US9304812B2 (en) 2016-04-05
GB201021414D0 (en) 2011-01-26
US11947999B2 (en) 2024-04-02
CN103314357A (en) 2013-09-18

Similar Documents

Publication Publication Date Title
US11947999B2 (en) Multi-phased and multi-threaded program execution based on SIMD ratio
US7269712B2 (en) Thread selection for fetching instructions for pipeline multi-threaded processor
US7577826B2 (en) Stall prediction thread management
US7659898B2 (en) Multi-execution resource graphics processor
US9811340B2 (en) Method and apparatus for reconstructing real program order of instructions in multi-strand out-of-order processor
JP5411587B2 (en) Multi-thread execution device and multi-thread execution method
US20080244247A1 (en) Processing long-latency instructions in a pipelined processor
JP5173714B2 (en) Multi-thread processor and interrupt processing method thereof
US20080163230A1 (en) Method and apparatus for selection among multiple execution threads
US20150074353A1 (en) System and Method for an Asynchronous Processor with Multiple Threading
US20110004881A1 (en) Look-ahead task management
EP1416377B1 (en) Processor system with a plurality of processor cores for executing tasks sequentially or in parallel
KR20150019349A (en) Multiple threads execution processor and its operating method
Bocci et al. Bringing heterogeneity to the CMS software framework
US20170139751A1 (en) Scheduling method and processing device using the same
Liu et al. Supporting soft real-time parallel applications on multicore processors
CN115617499A (en) System and method for GPU multi-core hyper-threading technology
US20030046517A1 (en) Apparatus to facilitate multithreading in a computer processor pipeline
CN116414464A (en) Method and device for scheduling tasks, electronic equipment and computer readable medium
CN111258655A (en) Fusion calculation method and readable storage medium
CN110088730B (en) Task processing method, device, medium and equipment
Bauer et al. PATS: a performance aware task scheduler for runtime reconfigurable processors
JP5573038B2 (en) Multi-thread processor and program
US8046569B2 (en) Processing element having dual control stores to minimize branch latency
Grudnitsky et al. MORP: Makespan optimization for processors with an embedded reconfigurable fabric