NASP/cuda-samples
35
0
Fork 0
mirror of https://github.com/NVIDIA/cuda-samples.git synced 2024-11-24 19:29:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

--Update README.md for cuda 10.1 release

Browse Source 
-- fix multidevice CG sample to check for concurrentManagedAccess as it needs systemwide atomics
This commit is contained in:
Mahesh Doijade 2019-01-23 16:56:20 +05:30
parent b458dafcd6
commit 1abc294982
2 changed files with 29 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

40
README.md
View File

@ -1,11 +1,19 @@
# CUDA Samples # CUDA Samples
Samples for CUDA Developers which demonstrates features in CUDA Toolkit. This version supports [CUDA Toolkit 10.0](https://developer.nvidia.com/cuda-downloads). Samples for CUDA Developers which demonstrates features in CUDA Toolkit. This version supports [CUDA Toolkit 10.1](https://developer.nvidia.com/cuda-downloads).
## Release Notes ## Release Notes
This section describes the release notes for the CUDA Samples on GitHub only. This section describes the release notes for the CUDA Samples on GitHub only.
### CUDA 10.1
* Added `immaTensorCoreGemm`. Demonstrates integer GEMM computation using the Warp Matrix Multiply and Accumulate (WMMA) API for integers employing the Tensor Cores.
* Added `simpleIPC`. Demonstrates Inter Process Communication with one process per GPU for computation.
* Added `nvJPEG`. Demonstrates single and batched decoding of jpeg images using NVJPEG Library.
* Added `bandwidthTest`. It measures the memcopy bandwidth of the GPU and memcpy bandwidth across PCI-e.
* Added `reduction`. Demonstrates several important optimization strategies for Data-Parallel Algorithms like reduction.
* Update all the samples to support CUDA 10.1.
### CUDA 10.0 ### CUDA 10.0
* Added `simpleCudaGraphs`. Demonstrates CUDA Graphs creation, instantiation and launch using Graphs APIs and Stream Capture APIs. * Added `simpleCudaGraphs`. Demonstrates CUDA Graphs creation, instantiation and launch using Graphs APIs and Stream Capture APIs.
* Added `conjugateGradientCudaGraphs`. Demonstrates conjugate gradient solver on GPU using CUBLAS and CUSPARSE library calls captured and called using CUDA Graph APIs. * Added `conjugateGradientCudaGraphs`. Demonstrates conjugate gradient solver on GPU using CUBLAS and CUSPARSE library calls captured and called using CUDA Graph APIs.
@ -38,7 +46,7 @@ This is the first release of CUDA Samples on GitHub:
### Prerequisites ### Prerequisites
Download and install the [CUDA Toolkit 10.0](https://developer.nvidia.com/cuda-downloads) for your corresponding platform. Download and install the [CUDA Toolkit 10.1](https://developer.nvidia.com/cuda-downloads) for your corresponding platform.
For system requirements and installation instructions of cuda toolkit, please refer to the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/), the [Windows Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/index.html), and the [Mac Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-mac-os-x/index.html). For system requirements and installation instructions of cuda toolkit, please refer to the [Linux Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-linux/), the [Windows Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/index.html), and the [Mac Installation Guide](http://docs.nvidia.com/cuda/cuda-installation-guide-mac-os-x/index.html).
### Getting the CUDA Samples ### Getting the CUDA Samples
@ -120,27 +128,29 @@ The samples makefiles can take advantage of certain options:
### Samples by OS ### Samples by OS
#### Linux #### Linux
**[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[conjugateGradientMultiBlockCG](./Samples/conjugateGradientMultiBlockCG)** | **[shfl_scan](./Samples/shfl_scan)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[conjugateGradientMultiBlockCG](./Samples/conjugateGradientMultiBlockCG)** | **[simpleIPC](./Samples/simpleIPC)** | **[shfl_scan](./Samples/shfl_scan)** |
---|---|---|---| ---|---|---|---|
**[simpleCudaGraphs](./Samples/simpleCudaGraphs)** | **[deviceQuery](./Samples/deviceQuery)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[nvJPEG](./Samples/nvJPEG)** | **[simpleCudaGraphs](./Samples/simpleCudaGraphs)** | **[deviceQuery](./Samples/deviceQuery)** |
**[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** | **[conjugateGradientMultiDeviceCG](./Samples/conjugateGradientMultiDeviceCG)** | **[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleCUFFT](./Samples/simpleCUFFT)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** | **[conjugateGradientMultiDeviceCG](./Samples/conjugateGradientMultiDeviceCG)** |
**[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[cudaTensorCoreGemm](./Samples/cudaTensorCoreGemm)** | **[simpleVulkan](./Samples/simpleVulkan)** | **[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleCUFFT](./Samples/simpleCUFFT)** | **[reduction](./Samples/reduction)** | **[immaTensorCoreGemm](./Samples/immaTensorCoreGemm)** |
**[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** | **[systemWideAtomics](./Samples/systemWideAtomics)** | **[bandwidthTest](./Samples/bandwidthTest)** | **[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[cudaTensorCoreGemm](./Samples/cudaTensorCoreGemm)** | **[simpleVulkan](./Samples/simpleVulkan)** |
**[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** | **[systemWideAtomics](./Samples/systemWideAtomics)** |
#### Windows #### Windows
**[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[conjugateGradientMultiBlockCG](./Samples/conjugateGradientMultiBlockCG)** | **[shfl_scan](./Samples/shfl_scan)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[conjugateGradientMultiBlockCG](./Samples/conjugateGradientMultiBlockCG)** | **[simpleIPC](./Samples/simpleIPC)** | **[shfl_scan](./Samples/shfl_scan)** |
---|---|---|---| ---|---|---|---|
**[simpleD3D12](./Samples/simpleD3D12)** | **[simpleCudaGraphs](./Samples/simpleCudaGraphs)** | **[deviceQuery](./Samples/deviceQuery)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[simpleD3D12](./Samples/simpleD3D12)** | **[simpleCudaGraphs](./Samples/simpleCudaGraphs)** | **[deviceQuery](./Samples/deviceQuery)** |
**[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** | **[conjugateGradientMultiDeviceCG](./Samples/conjugateGradientMultiDeviceCG)** | **[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** | **[conjugateGradientMultiDeviceCG](./Samples/conjugateGradientMultiDeviceCG)** |
**[simpleCUFFT](./Samples/simpleCUFFT)** | **[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[cudaTensorCoreGemm](./Samples/cudaTensorCoreGemm)** | **[simpleVulkan](./Samples/simpleVulkan)** | **[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleCUFFT](./Samples/simpleCUFFT)** | **[reduction](./Samples/reduction)** | **[immaTensorCoreGemm](./Samples/immaTensorCoreGemm)** |
**[bandwidthTest](./Samples/bandwidthTest)** | **[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[cudaTensorCoreGemm](./Samples/cudaTensorCoreGemm)** | **[simpleVulkan](./Samples/simpleVulkan)** |
**[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** | **[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** |
#### Mac OSX #### Mac OSX
**[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[shfl_scan](./Samples/shfl_scan)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[simpleCudaGraphs](./Samples/simpleCudaGraphs)** | **[warpAggregatedAtomicsCG](./Samples/warpAggregatedAtomicsCG)** | **[shfl_scan](./Samples/shfl_scan)** | **[conjugateGradientCudaGraphs](./Samples/conjugateGradientCudaGraphs)** | **[simpleCudaGraphs](./Samples/simpleCudaGraphs)** |
---|---|---|---| ---|---|---|---|
**[deviceQuery](./Samples/deviceQuery)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** | **[deviceQuery](./Samples/deviceQuery)** | **[simpleVoteIntrinsics](./Samples/simpleVoteIntrinsics)** | **[simpleCUBLASXT](./Samples/simpleCUBLASXT)** | **[UnifiedMemoryPerf](./Samples/UnifiedMemoryPerf)** |
**[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleCUFFT](./Samples/simpleCUFFT)** | **[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[matrixMulDrv](./Samples/matrixMulDrv)** | **[simpleCUFFT](./Samples/simpleCUFFT)** | **[reduction](./Samples/reduction)** | **[bandwidthTest](./Samples/bandwidthTest)** |
**[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** | **[simpleCUBLAS](./Samples/simpleCUBLAS)** | **[vectorAdd_nvrtc](./Samples/vectorAdd_nvrtc)** | **[p2pBandwidthLatencyTest](./Samples/p2pBandwidthLatencyTest)** | **[matrixMul](./Samples/matrixMul)** |
## Dependencies ## Dependencies
@ -262,6 +272,10 @@ NPP (NVIDIA Performance Primitives) provides GPU-accelerated image, video, and s
NVGRAPH is a GPU-accelerated graph analytics library. NVGRAPH is a GPU-accelerated graph analytics library.
#### NVJPEG
NVJPEG library provides high-performance, GPU accelerated JPEG decoding functionality for image formats commonly used in deep learning and hyperscale multimedia applications.
#### NVRTC #### NVRTC
NVRTC (CUDA RunTime Compilation) is a runtime compilation library for CUDA C++. NVRTC (CUDA RunTime Compilation) is a runtime compilation library for CUDA C++.

4
Samples/conjugateGradientMultiDeviceCG/conjugateGradientMultiDeviceCG.cu
View File

@ -415,7 +415,7 @@ void getIdenticalGPUs(int num_of_gpus, std::set<int> &identicalGPUs) {
identicalGPUs.erase(it); identicalGPUs.erase(it);
} }
if (!deviceProp.cooperativeMultiDeviceLaunch || if (!deviceProp.cooperativeMultiDeviceLaunch ||
!deviceProp.managedMemory) { !deviceProp.concurrentManagedAccess) {
identicalGPUs.erase(it); identicalGPUs.erase(it);
} }
it++; it++;
@ -450,7 +450,7 @@ int main(int argc, char **argv) {
if (identicalGPUs.size() <= 1) { if (identicalGPUs.size() <= 1) {
printf( printf(
"No Two or more GPUs with same architecture capable of " "No Two or more GPUs with same architecture capable of "
"cooperativeMultiDeviceLaunch & managedMemory found. \nWaiving the sample\n"); "cooperativeMultiDeviceLaunch & concurrentManagedAccess found. \nWaiving the sample\n");
exit(EXIT_WAIVED); exit(EXIT_WAIVED);
} }