Nvidia Cuda Arc

PowerPoint Presentation

CUDA ARCHITECTURE

Prepared By:Shubham Agrawal(120750107060)Jainam Jain(120750107036)Dwij Erda(120750107025)Affiliated To:SHANKERSINH VAGHELA BAPU INSTITUTE OF TECHNOLOGY

INTRODUCTIONCUDA(Compute Unified Device Architecture) is aparallel computingplatform and programming model created byNVIDIAand implemented by thegraphics processing units(GPUs) that they produce.CUDA gives program developers direct access to the virtualinstruction setand memory of the parallel computational elements in CUDA GPUs.Using CUDA, the GPUs can be used for general purpose processing (i.e., not exclusively graphics); this approach is known asGPGPU. Unlike CPUs, however, GPUs have a parallel throughput architecture that emphasizes executing many concurrent threads slowly, rather than executing a single thread very quickly.

The CUDA platform is accessible to software developers through CUDA-accelerated libraries,compiler directives and extensions to industry-standard programming languages, includingC,C++andFortran.In thecomputer gameindustry, GPUs are used not only for graphics rendering but also ingame physics calculations(physical effects like debris, smoke, fire, fluids) examples includePhysXandBullet. CUDA has also been used to accelerate non-graphical applications incomputational biology,cryptographyand other fields by anorder of magnitudeor more.CUDA provides both a low levelAPI(Application Programming Interface)and a higher level API.

Concept of CUDA(GPGPU) Latency Processor + Throughput Processor

HISTORYThe initial CUDASDKwas made public on 15 February 2007, forMicrosoft Windows andLinux.Mac OS Xsupport was later added in version 2.0,which supersedes the beta released February 14, 2008.CUDA works with all Nvidia GPUs from the G8x series onwards, includingGeForce,Quadroand theTeslaline. CUDA is compatible with most standard operating systems. Nvidia states that programs developed for the G8x series will also work without modification on all future Nvidia video cards, due to binary compatibility.

EVALUATIONNvidia TeslaTesla K40Tesla K20XTesla K20Tesla K10Tesla C2050/2070Tesla M2050/M2070Tesla S2050Tesla S1070Tesla M1060Tesla C1060Tesla C870Tesla D870Tesla S870

Nvidia QuadroQuadro K6000Quadro K5000Quadro K4000Quadro K2000DQuadro K2000Quadro K600Quadro 6000Quadro 5000Quadro 4000Quadro 2000Quadro 600Quadro FX 5800Quadro FX 5600Quadro FX 4800

Nvidia GeForce MobileGeForce GTX 880MGeForce GTX 870MGeForce GTX 860MGeForce GTX 850MGeForce 845MGeForce 840MGeForce 830MGeForce GTX 780MGeForce GTX 770MGeForce GTX 765MGeForce GTX 760MGeForce GT 750MGeForce GT 745MGeForce GT 740M

Nvidia GeForceGeForce GTX Titan ZGeForce GTX TITAN BlackGeForce GTX TITANGeForce GTX 780 TiGeForce GTX 780GeForce GTX 770GeForce GTX 760GeForce GTX 750 TiGeForce GTX 750GeForce GT 740GeForce GT 730GeForce GTX 690GeForce GTX 680GeForce GTX 670

PROCESSING FLOW

GPU ARCHITECTURETwo Main ComponentsGlobal memoryAnalogous to RAM in a CPU serverAccessible by both GPU and CPUCurrently up to 6 GBBandwidth currently up to 150 GB/s for Quadro and Tesla productsECC on/off option for Quadro and Tesla productsStreaming Multiprocessors (SMs)Perform the actual computationsEach SM has its own:Control units, registers, execution pipelines, caches

GPU ARCHITECTUREFermi: Streaming Multiprocessor (SM)32 CUDA Cores per SM32 fp32 ops/clock16 fp64 ops/clock32 int32 ops/clock2 warp schedulersUp to 1536 threads concurrently4 special-function units64KB shared mem+ L1 cache32K 32-bit registersRegister

GPU ARCHITECTUREFermi: CUDA CoreFloating point & Integer unitIEEE 754-2008 floating-point standardFused multiply-add (FMA) instruction for both single and double precisionLogic unitMove, compare unitBranch unitRegister

Recent ApplicationsGEFORCE GTX TITAN Z

NVIDIA TABLET

NVIDIA SHIELD

NVIDIA SHADOWPLAY

REFERENCESWe are very thankful to:GOOGLE-For Images and gadgetsWIKIPEDIA-For informationYOU TUBE-For videoAnd the most important one the official site of nvidiawww.nvidia.in

THANK YOU