10.07.2015 Views

ATI Stream Computing OpenCL Programming Guide - CiteSeerX

ATI Stream Computing OpenCL Programming Guide - CiteSeerX

ATI Stream Computing OpenCL Programming Guide - CiteSeerX

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

<strong>ATI</strong> STREAM COMPUTINGCompute DevicePrivate Memory(Reg Files)Private Memory(Reg Files) m12. . .Compute Unit nPrivate Memory(Reg Files) 1. . .Private Memory(Reg Files) mProc. Elem.(ALU)Proc. Elem.(ALU)Local Mem.(LDS) 12R/OL1W/OCompute Unit 1atomicR/WColor Buffer. . .ALUProc. Elem.(ALU)Local Mem.(LDS) nProc. Elem.(ALU)read onlyL1Color Bufferwrite onlyatomic read / writeIMAGE / CONSTANT DATA CACHE (L2)(read-only)Compute DeviceMemory(VRAM)GLOBAL MEMORYCONSTANT MEMORYDMAHostPCIeFigure 1.7Interrelationship of Memory DomainsFigure 1.8 illustrates the standard dataflow between host (CPU) and GPU.HOSTPCIeGLOBALLOCALPRIVATEFigure 1.8Dataflow between Host and GPUThere are two ways to copy data from the host to the GPU compute devicememory:• Implicitly by using clEnqueueMapBuffer and clEnqueueUnMapMemObject.• Explicitly through clEnqueueReadBuffer and clEnqueueWriteBuffer(clEnqueueReadImage, clEnqueueWriteImage.).1.4 Memory Architecture and Access 1-11Copyright © 2010 Advanced Micro Devices, Inc. All rights reserved.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!