《内存计算PPT.ppt》由会员分享，可在线阅读，更多相关《内存计算PPT.ppt（23页珍藏版）》请在三一办公上搜索。
1、In-memory computing,A solution to the Von Neumann bottleneck,Sylvain EUDIERUnion College,2004 MSCS Candidate,Seminar-Processing in Memory,Plan,Introduction to a new architectureDifferent architecturesThe C-RAM Architecture:Implication/ApplicationPerformancesConclusion,Seminar-Processing in Memory,In
2、troduction,Von Neumann architectureThe situation(gap evolution)Some improvements were madeCan we avoid this bottleneck?(graph),Seminar-Processing in Memory,Different Architectures,Architectures and designsIRAM(design)RAW(design)CRAM(design),Seminar-Processing in Memory,The C-RAM(Computational RAM)ar
3、chitecture,ApplicationsPerformancesImplicationsNew software designEnergy consumption,Seminar-Processing in Memory,CRAM Applications,Image processing:Low-level adjustments(brightness,average filter)Databases searches:Equivalence,Extremes,between limitsMultimedia Compression:MPEG Motion estimation,Sem
4、inar-Processing in Memory,Performances-Configs,CRAM 200Mhz;32MB;64K PEs on a Pentium 133Mhz(simulated)Pentium 133Mhz with 32 MB RamSun SPARC Station 167 Mhz CPU with 64 MB,Seminar-Processing in Memory,Performances Basic ops,Ops complexity,Seminar-Processing in Memory,Performances-Comparison,Seminar-
5、Processing in Memory,New Software Design(Step 1),Think Parallel(pseudo code),Seminar-Processing in Memory,New Software Design(Step 2),Use a different language(modified C+),Seminar-Processing in Memory,New Software Design(Step 3),Possibly coding in assembly to optimize,Seminar-Processing in Memory,CR
6、AM Energy Consumption,We avoid the use of a busWe have a direct access to memoryNo overhead in communicationFinally the CRAM use 20 times less energy therefore less heat,Seminar-Processing in Memory,The future,Which architecture will be chosen?End of todays architecture?A PetaOps is feasible with CR
7、AMBlue Gene/P aims at the petaFlops(view),Seminar-Processing in Memory,Questions?,Seminar-Processing in Memory,Memory bandwidth in a computer,Back,Seminar-Processing in Memory,IRAM Design,Back,Seminar-Processing in Memory,RAW Design,Back,Seminar-Processing in Memory,CRAM Design,Back,Seminar-Processing in Memory,Operations Complexity for CRAM,Back,Seminar-Processing in Memory,Blue Gene/P,Scale,Seminar-Processing in Memory,Computing power Scale,Back,Seminar-Processing in Memory,Memory processors gap evolution,Back,
