mirror of
https://github.com/NVIDIA/cuda-samples.git
synced 2024-10-06 18:19:15 +08:00
100 lines
2.8 KiB
XML
100 lines
2.8 KiB
XML
<?xml version="1.0" encoding="UTF-8"?>
|
|
<!DOCTYPE entry SYSTEM "SamplesInfo.dtd">
|
|
<entry>
|
|
<name>globalToShmemAsyncCopy</name>
|
|
<cflags>
|
|
<flag>--std=c++11</flag>
|
|
</cflags>
|
|
<cuda_api_list>
|
|
<toolkit>cudaStreamCreateWithFlags</toolkit>
|
|
<toolkit>cudaMalloc</toolkit>
|
|
<toolkit>cudaDeviceGetAttribute</toolkit>
|
|
<toolkit>cudaFree</toolkit>
|
|
<toolkit>cudaMallocHost</toolkit>
|
|
<toolkit>cudaEventSynchronize</toolkit>
|
|
<toolkit>cudaEventRecord</toolkit>
|
|
<toolkit>cudaFreeHost</toolkit>
|
|
<toolkit>cudaStreamSynchronize</toolkit>
|
|
<toolkit>cudaEventDestroy</toolkit>
|
|
<toolkit>cudaEventElapsedTime</toolkit>
|
|
<toolkit>cudaMemsetAsync</toolkit>
|
|
<toolkit>cudaMemcpyAsync</toolkit>
|
|
<toolkit>cudaEventCreate</toolkit>
|
|
</cuda_api_list>
|
|
<description><![CDATA[This sample implements matrix multiplication which uses asynchronous copy of data from global to shared memory when on compute capability 8.0 or higher. Also demonstrates arrive-wait barrier for synchronization.]]></description>
|
|
<devicecompilation>whole</devicecompilation>
|
|
<includepaths>
|
|
<path>./</path>
|
|
<path>../</path>
|
|
<path>../../../Common</path>
|
|
</includepaths>
|
|
<keyconcepts>
|
|
<concept level="basic">CUDA Runtime API</concept>
|
|
<concept level="basic">Linear Algebra</concept>
|
|
<concept level="basic">CPP11 CUDA</concept>
|
|
</keyconcepts>
|
|
<keywords>
|
|
<keyword>CUDA</keyword>
|
|
<keyword>matrix multiply</keyword>
|
|
<keyword>Async copy</keyword>
|
|
<keyword>CPP11</keyword>
|
|
<keyword>GCC 5.1.0</keyword>
|
|
</keywords>
|
|
<libraries>
|
|
</libraries>
|
|
<librarypaths>
|
|
</librarypaths>
|
|
<nsight_eclipse>true</nsight_eclipse>
|
|
<primary_file>globalToShmemAsyncCopy.cu</primary_file>
|
|
<required_dependencies>
|
|
<dependency>CPP11</dependency>
|
|
</required_dependencies>
|
|
<scopes>
|
|
<scope>1:CUDA Basic Topics</scope>
|
|
<scope>3:Linear Algebra</scope>
|
|
</scopes>
|
|
<sm-arch>sm70</sm-arch>
|
|
<sm-arch>sm72</sm-arch>
|
|
<sm-arch>sm75</sm-arch>
|
|
<sm-arch>sm80</sm-arch>
|
|
<sm-arch>sm86</sm-arch>
|
|
<sm-arch>sm87</sm-arch>
|
|
<sm-arch>sm89</sm-arch>
|
|
<sm-arch>sm90</sm-arch>
|
|
<supported_envs>
|
|
<env>
|
|
<arch>x86_64</arch>
|
|
<platform>linux</platform>
|
|
</env>
|
|
<env>
|
|
<arch>x86_64</arch>
|
|
<platform>macosx</platform>
|
|
</env>
|
|
<env>
|
|
<arch>arm</arch>
|
|
</env>
|
|
<env>
|
|
<arch>sbsa</arch>
|
|
</env>
|
|
<env>
|
|
<arch>ppc64le</arch>
|
|
<platform>linux</platform>
|
|
</env>
|
|
<env>
|
|
<arch>aarch64</arch>
|
|
<platform>linux</platform>
|
|
</env>
|
|
<env>
|
|
<arch>aarch64</arch>
|
|
<platform>qnx</platform>
|
|
</env>
|
|
<env>
|
|
<platform>windows7</platform>
|
|
</env>
|
|
</supported_envs>
|
|
<supported_sm_architectures>
|
|
<from>7.0</from>
|
|
</supported_sm_architectures>
|
|
<title>Global Memory to Shared Memory Async Copy</title>
|
|
</entry>
|