有多少个“CUDA 核心”? GPU 的每个多处理器都有吗?
我知道 Fermi 架构之前的设备在单个多处理器中具有 8 个 SP。费米架构中的计数相同吗?
I know that devices before the Fermi architecture had 8 SPs in a single multiprocessor. Is the count same in Fermi architecture?
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(3)
答案取决于 CUDA 设备的计算能力属性。这些数字是:
请参阅CUDA C 编程指南<的附录 G< /a>.
The answer depends on the Compute Capability property of the CUDA device. The numbers are:
See appendix G of the CUDA C Programming Guide.
通过执行DeviceQuery.exe可以找到多处理器(MP)的数量和每个MP的核心的数量。它位于 GPU 计算 SDK 安装的
%NVSDKCOMPUTE_ROOT%/C/bin
目录中。查看 DeviceQuery 的代码(位于
%NVSDKCOMPUTE_ROOT%/C/src/DeviceQuery
),可以发现核心数量是通过传递 xy CUDA Capability 数字来计算的到 ConvertSMVer2Cores 实用函数。从ConvertSMVer2Cores的代码中可以看出能力与核心数之间的关系:
The number of Multiprocessors (MP) and the number of cores per MP can be found by executing DeviceQuery.exe. It is found in the
%NVSDKCOMPUTE_ROOT%/C/bin
directory of the GPU Computing SDK installation.A look at the code of DeviceQuery (found in
%NVSDKCOMPUTE_ROOT%/C/src/DeviceQuery
) reveals that it the number of cores is calculated by passing the x.y CUDA Capability numbers to the ConvertSMVer2Cores utility function.From the code of ConvertSMVer2Cores this relationship between the capability and core count can be seen:
更新 @AshwinNanjappa 对 CUDA 7.5 的回答:
注意:
$CUDA_SAMPLES_DIR/common/inc/helper_cuda.h
获得了附加信息。Update of @AshwinNanjappa's answer for CUDA 7.5:
Notes:
$CUDA_SAMPLES_DIR/common/inc/helper_cuda.h
.