You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Both CUDA and OpenCL allow you to use PTX as the source for a kernel, albeit somewhat differently than CUDA C++ or OpenCL C. The kernel runner should be adapted to accept PTX as input, and complete the compilation into a directly executable kernel. At least initially, we can require the user to tell us explicitly that the kernel is PTX (in addition to saying whether it's CUDA or OpenCL code), so that we don't need to worry about source file type autodetection.
The text was updated successfully, but these errors were encountered:
Both CUDA and OpenCL allow you to use PTX as the source for a kernel, albeit somewhat differently than CUDA C++ or OpenCL C. The kernel runner should be adapted to accept PTX as input, and complete the compilation into a directly executable kernel. At least initially, we can require the user to tell us explicitly that the kernel is PTX (in addition to saying whether it's CUDA or OpenCL code), so that we don't need to worry about source file type autodetection.
The text was updated successfully, but these errors were encountered: