CUDA_C_Programming_Guide7.5资源-CSDN文库

cuda7.5,

cuda编程指南

5星 · 超过95%的资源需积分: 9 12 浏览量 2015-11-16 12:33:26 上传评论 1 收藏 3.26MB PDF 举报

资源推荐

资源详情

资源评论

CUDA C PROGRAMMING GUIDE

PG-02829-001_v7.5 | September 2015

Design Guide

www.nvidia.com

CUDA C Programming Guide PG-02829-001_v7.5|iii

TABLE OF CONTENTS

Chapter1.Introduction.........................................................................................1

1.1.From Graphics Processing to General Purpose Parallel Computing............................... 1

1.2.CUDA

: A General-Purpose Parallel Computing Platform and Programming Model.............4

1.3.A Scalable Programming Model.........................................................................5

1.4.Document Structure...................................................................................... 7

Chapter2.Programming Model............................................................................... 9

2.1. Kernels......................................................................................................9

2.2.Thread Hierarchy........................................................................................ 10

2.3.Memory Hierarchy....................................................................................... 12

2.4.Heterogeneous Programming.......................................................................... 14

2.5.Compute Capability..................................................................................... 16

Chapter3.Programming Interface..........................................................................17

3.1.Compilation with NVCC................................................................................ 17

3.1.1.Compilation Workflow.............................................................................18

3.1.1.1.Offline Compilation.......................................................................... 18

3.1.1.2.Just-in-Time Compilation....................................................................18

3.1.2.Binary Compatibility...............................................................................18

3.1.3.PTX Compatibility..................................................................................19

3.1.4.Application Compatibility.........................................................................19

3.1.5.C/C++ Compatibility............................................................................... 20

3.1.6.64-Bit Compatibility............................................................................... 20

3.2.CUDA C Runtime.........................................................................................20

3.2.1. Initialization.........................................................................................21

3.2.2.Device Memory..................................................................................... 21

3.2.3.Shared Memory..................................................................................... 24

3.2.4.Page-Locked Host Memory........................................................................29

3.2.4.1.Portable Memory..............................................................................30

3.2.4.2.Write-Combining Memory....................................................................30

3.2.4.3.Mapped Memory...............................................................................30

3.2.5.Asynchronous Concurrent Execution............................................................ 31

3.2.5.1.Concurrent Execution between Host and Device........................................32

3.2.5.2.Concurrent Kernel Execution............................................................... 32

3.2.5.3.Overlap of Data Transfer and Kernel Execution......................................... 32

3.2.5.4.Concurrent Data Transfers.................................................................. 33

3.2.5.5. Streams......................................................................................... 33

3.2.5.6. Events...........................................................................................37

3.2.5.7.Synchronous Calls.............................................................................37

3.2.6.Multi-Device System............................................................................... 38

3.2.6.1.Device Enumeration.......................................................................... 38

3.2.6.2.Device Selection.............................................................................. 38

www.nvidia.com

CUDA C Programming Guide PG-02829-001_v7.5|iv

3.2.6.3.Stream and Event Behavior................................................................. 38

3.2.6.4.Peer-to-Peer Memory Access................................................................39

3.2.6.5.Peer-to-Peer Memory Copy..................................................................39

3.2.7.Unified Virtual Address Space................................................................... 40

3.2.8.Interprocess Communication..................................................................... 41

3.2.9.Error Checking......................................................................................41

3.2.10. Call Stack.......................................................................................... 42

3.2.11.Texture and Surface Memory................................................................... 42

3.2.11.1.Texture Memory............................................................................. 42

3.2.11.2.Surface Memory............................................................................. 52

3.2.11.3.CUDA Arrays..................................................................................56

3.2.11.4.Read/Write Coherency..................................................................... 56

3.2.12.Graphics Interoperability........................................................................56

3.2.12.1.OpenGL Interoperability................................................................... 57

3.2.12.2.Direct3D Interoperability...................................................................59

3.2.12.3.SLI Interoperability..........................................................................65

3.3.Versioning and Compatibility.......................................................................... 66

3.4.Compute Modes..........................................................................................67

3.5. Mode Switches........................................................................................... 68

3.6.Tesla Compute Cluster Mode for Windows.......................................................... 68

Chapter4.Hardware Implementation......................................................................69

4.1.SIMT Architecture....................................................................................... 69

4.2.Hardware Multithreading...............................................................................71

Chapter5.Performance Guidelines........................................................................ 72

5.1.Overall Performance Optimization Strategies...................................................... 72

5.2.Maximize Utilization.................................................................................... 72

5.2.1.Application Level...................................................................................72

5.2.2. Device Level........................................................................................ 73

5.2.3.Multiprocessor Level...............................................................................73

5.2.3.1.Occupancy Calculator........................................................................ 75

5.3.Maximize Memory Throughput........................................................................ 77

5.3.1.Data Transfer between Host and Device....................................................... 78

5.3.2.Device Memory Accesses..........................................................................79

5.4.Maximize Instruction Throughput.....................................................................83

5.4.1.Arithmetic Instructions............................................................................83

5.4.2.Control Flow Instructions......................................................................... 87

5.4.3.Synchronization Instruction.......................................................................88

AppendixA.CUDA-Enabled GPUs........................................................................... 89

AppendixB.C Language Extensions........................................................................ 90

B.1.Function Type Qualifiers............................................................................... 90

B.1.1. __device__.......................................................................................... 90

B.1.2. __global__...........................................................................................90

B.1.3. __host__............................................................................................. 90

www.nvidia.com

CUDA C Programming Guide PG-02829-001_v7.5|v

B.1.4.__noinline__ and __forceinline__............................................................... 91

B.2.Variable Type Qualifiers................................................................................91

B.2.1. __device__.......................................................................................... 91

B.2.2.__constant__........................................................................................92

B.2.3. __shared__.......................................................................................... 92

B.2.4.__managed__....................................................................................... 93

B.2.5. __restrict__......................................................................................... 93

B.3.Built-in Vector Types................................................................................... 94

B.3.1.char, short, int, long, longlong, float, double................................................ 94

B.3.2. dim3.................................................................................................. 95

B.4.Built-in Variables........................................................................................ 96

B.4.1. gridDim.............................................................................................. 96

B.4.2. blockIdx..............................................................................................96

B.4.3. blockDim.............................................................................................96

B.4.4. threadIdx............................................................................................ 96

B.4.5. warpSize............................................................................................. 96

B.5.Memory Fence Functions...............................................................................96

B.6.Synchronization Functions............................................................................. 99

B.7.Mathematical Functions...............................................................................100

B.8.Texture Functions...................................................................................... 100

B.8.1.Texture Object API............................................................................... 101

B.8.1.1.tex1Dfetch()..................................................................................101

B.8.1.2. tex1D()........................................................................................ 101

B.8.1.3.tex1DLod()....................................................................................101

B.8.1.4.tex1DGrad().................................................................................. 101

B.8.1.5. tex2D()........................................................................................ 101

B.8.1.6.tex2DLod()....................................................................................101

B.8.1.7.tex2DGrad().................................................................................. 102

B.8.1.8. tex3D()........................................................................................ 102

B.8.1.9.tex3DLod()....................................................................................102

B.8.1.10.tex3DGrad().................................................................................102

B.8.1.11.tex1DLayered()............................................................................. 102

B.8.1.12.tex1DLayeredLod().........................................................................102

B.8.1.13.tex1DLayeredGrad()....................................................................... 103

B.8.1.14.tex2DLayered()............................................................................. 103

B.8.1.15.tex2DLayeredLod().........................................................................103

B.8.1.16.tex2DLayeredGrad()....................................................................... 103

B.8.1.17.texCubemap().............................................................................. 103

B.8.1.18.texCubemapLod().......................................................................... 103

B.8.1.19.texCubemapLayered().....................................................................104

B.8.1.20.texCubemapLayeredLod()................................................................ 104

B.8.1.21.tex2Dgather()...............................................................................104

B.8.2.Texture Reference API........................................................................... 105

剩余260页未读，继续阅读

评论收藏

内容反馈

海角儿

2016-07-14

真实可靠,赞
放屁带出翔丶

2018-09-19

书很不错，工具书

yang_88

粉丝: 0
资源: 6

CUDA_C_Programming_Guide 7.5

最新资源

CUDA_C_Programming_Guide 7.5

CUDA_C_Programming_Guide

CUDA C Programming Guide v8.0

CUDA C Programming Guide v9.0

CUDA_C_Programming_Guide.pdf

CUDA_C_Programming_Guide_CN

cuda8.0_cuda_c_programming_guide_2017version

CUDA_C_Programming_Guide v10.0.pdf

CUDA_C_Programming_Guide _9.0

CUDA_2.0编程指南_NVIDIA_CUDA_Programming_Guide_2.0Final

CUDA_C_Best_Practices_Guide_cuda_GPU_

cuda_12.5.0_555.85_windows.exe.012

NVIDIA_CUDA_C_ProgrammingGuide（CUDA 编程指南）

NVIDIA CUDA Programming Guide

CUDA_C_Programming_Guide 4.1

cuda_12.5.0_555.85_windows.exe.004

cuda_11.1.0_455.23.05_linux.tar.gz3

NAVID CUDA_C_Programming_Guide

最新资源