Unformatted text preview:

The IBM Cell ArchitectureTopicsSlide 3MotivationGoalsSlide 6Software Cells: The ConceptSoftware Cells: FormattingComparison with Dataflow ArchitectureSlide 10Machine ArchitectureSoC Architecture(Envisioned) SPU ArchitectureSlide 14Prototype Chip FloorplanNotes on PrototypeSlide 17Programmer’s Interface: Two-PartsSlide 19Cell ReferencesSam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureThe IBM Cell ArchitectureThe IBM Cell ArchitectureSam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureApril 18, 2006April 18, 2006Sam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureTopicsTopics1.1.OverviewOverview2.2.Software CellsSoftware Cells3.3.Machine ArchitectureMachine Architecture4.4.Product PrototypeProduct Prototype5.5.Programmer’s InterfaceProgrammer’s Interface6.6.References and GlossaryReferences and GlossarySam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureTopicsTopics1.1.OverviewOverview2.2.Software CellsSoftware Cells3.3.Machine ArchitectureMachine Architecture4.4.Product PrototypeProduct Prototype5.5.Programmer’s InterfaceProgrammer’s Interface6.6.References and GlossaryReferences and GlossarySam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureMotivationMotivationIBM’s formal name for Cell is “Cell Broadband Engine IBM’s formal name for Cell is “Cell Broadband Engine Architecture” (CBEA)Architecture” (CBEA)Sony wanted:Sony wanted:Quantum leap in performance over PlayStation 2’s “Emotion Engine” chip (made by Toshiba)Toshiba wanted:Toshiba wanted:Remain a part of volume manufacturing for Sony PlayStationIBM wanted:IBM wanted:A piece of the PlayStation 3 pieA second try at network processor architectureSomething reusable, applicable far beyond PlayStationSam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureGoalsGoalsApplication domainsApplication domainsGraphics Rendering ($$)DSP & Multimedia Processing ($$)CryptographyPhysics simulationsMatrix math and other scientific processingHeavy use of SIMD – why?Heavy use of SIMD – why?Cray and similar machines of 1970s achieved performance through vectorization rather than MIMD parallelizationThe above applications are areas in which SIMD is still the best architectureSam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureTopicsTopics1.1.OverviewOverview2.2.Software CellsSoftware Cells3.3.Machine ArchitectureMachine Architecture4.4.Product PrototypeProduct Prototype5.5.Programmer’s InterfaceProgrammer’s Interface6.6.References and GlossaryReferences and GlossarySam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureSoftware Cells: The ConceptSoftware Cells: The ConceptDefinitionDefinitionBundle of application code and working dataFeaturesFeaturesNecessarily object-orientedCells can migrate to any processor – local or remoteDistributed processing is native, and actually assumed•Execution of cell code actually looks like a remote procedure callA cell contains everything it needs to execute autonomously without references to other memory, programs or resourcesHighly secure model!Sam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureSoftware Software Cells: Cells: FormattingFormattingSource:U.S. Patent#6,809,734Sam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureComparison with Dataflow ArchitectureComparison with Dataflow ArchitectureGranularityGranularityDataflow execution granularity is one instructionCell execution granularity is a procedure, or several hundred instructionsopcodeoperand Aaddressoperand BaddressdestinationaddressDataflow instruction template:Sam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureTopicsTopics1.1.OverviewOverview2.2.Software CellsSoftware Cells3.3.Machine ArchitectureMachine Architecture4.4.Product PrototypeProduct Prototype5.5.Programmer’s InterfaceProgrammer’s Interface6.6.References and GlossaryReferences and GlossarySam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureMachine ArchitectureMachine ArchitectureEach Cell SoC contains:Each Cell SoC contains:Conventional processor (PPE), for control and a lightweight OS•2-way SMT, 2-way superscalar in-order Power coreMultiple Synergistic Processing Elements (SPEs)•These are execution engines for RPC of a software-cellDMA interface to memory and I/OElement Interconnect Bus (EIB), actually a ring busEach SPE contains:Each SPE contains:128 registers, 128 bits wide in unified regfile (2Kbytes of registers!)256 Kbytes local memory4 SIMD integer pipelines/ALUs4 SIMD floating point pipelines/FPUsSam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer ArchitectureSoC ArchitectureSoC ArchitectureALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemoryALUs (4)FPUs (4)regfile128x128256KBlocalmemory64-bit SMTPower core,2x in-ordersuperscalar512K L2I$ D$EIBDMA, I/OControllersPPESam SandboteSam SandboteCSE 8383 Advanced Computer ArchitectureCSE 8383 Advanced Computer Architecture(Envisioned) SPU Architecture(Envisioned) SPU ArchitectureResources for execution of multiple software cells are Resources for execution of multiple software cells are reserved in advance by the PPE:reserved in advance by the PPE:Some portion of local memoryOne or more dedicated integer/FP pipelinesNot SMT – pipelines are allocated in a dedicated way for the duration of the execution of the cellExecution is supposed to be entirely self-containedExecution is supposed to be entirely self-containedSoftware cell is small enough to execute on only one APUNo use of DRAM – the only addressable memory is local•Local memory is not cache – no


View Full Document

SMU CSE 8383 - The IBM Cell Architecture

Download The IBM Cell Architecture
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view The IBM Cell Architecture and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view The IBM Cell Architecture 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?