ComputeUnifiedDeviceArchitecture(CUDA)isNVIDIA'sGPUcomputingplatformandapplicationprogramminginterface.It'sdesignedtoworkwithprogramminglanguagessuchasC,C++,andPython.WithCUDA,youcanleverageaGPU'sparallelcomputingpowerforarangeofhigh-performancecomputingapplicationsinthefieldsofscience,healthcare,anddeeplearning.LearnCUDAProgrammingwillhelpyoulearnGPUparallelprogrammingandunderstanditsmodernapplications.Inthisbook,you'lldiscoverCUDAprogrammingapproachesformodernGPUarchitectures.You'llnotonlybeguidedthroughGPUfeatures,tools,andAPIs,you'llalsolearnhowtoanalyzeperformancewithsampleparallelprogrammingalgorithms.ThisbookwillhelpyouoptimizetheperformanceofyourappsbygivinginsightsintoCUDAprogrammingplatformswithvariouslibraries,compilerdirectives(OpenACC),andotherlanguages.Asyouprogress,you'lllearnhowadditionalcomputingpowercanbegeneratedusingmultipleGPUsinaboxorinmultipleboxes.Finally,you'llexplorehowCUDAacceleratesdeeplearningalgorithms,includingconvolutionalneuralnetworks(CNNs)andrecurrentneuralnetworks(RNNs).BytheendofthisCUDAbook,you'llbeequippedwiththeskillsyouneedtointegratethepowerofGPUcomputinginyourapplications.