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
I'm trying to run SYCL-based software on gpu (using dpctl ) and it requires the compute runtime to be installed to detect gpu devices.
The issue I have is that using ubuntu build the gpu devices are not detected. The same version downloaded from github works. But for end users it's much easier to use the build available in the official repo.
This could be an issue on the side of the ubuntu build, could it also be a bug in SYCL ?
edit 19/12: the dpcpp runtime must be pinned until IntelPython/dpctl#1022 is solved, replace intel::dpcpp_linux-64 with "intel::dpcpp_linux-64<2023.0.0' in the conda install commmand.
The text was updated successfully, but these errors were encountered:
fcharras
changed the title
Can't detect gpu with intel-compute-runtime from ubuntu reposiroties
Can't detect gpu with intel-compute-runtime from ubuntu repositories
Dec 6, 2022
Based on analysis/discussion happened in intel/compute-runtime#588, this seems to be a problem with Ubuntu package, and not with SYCL compiler and/or runtime hosted at intel/llvm. Closing this issue based on that. Please re-open or comment if you still have some unresolved concerns
Describe the bug
I'm trying to run SYCL-based software on gpu (using dpctl ) and it requires the compute runtime to be installed to detect gpu devices.
The issue I have is that using ubuntu build the gpu devices are not detected. The same version downloaded from github works. But for end users it's much easier to use the build available in the official repo.
This could be an issue on the side of the ubuntu build, could it also be a bug in SYCL ?
Mirror issues:
To Reproduce
Run this sequence of instructions from within a new ubuntu:jammy docker container:
The last command should list a gpu device but doesn't.
Replace with the build from intel/compute-runtime github releases:
The last command now lists a gpu device.
Environment (please complete the following information):
Running in a new ubuntu:jammy docker container.
Output from
mamba list --explicit
:edit 19/12: the dpcpp runtime must be pinned until IntelPython/dpctl#1022 is solved, replace
intel::dpcpp_linux-64
with"intel::dpcpp_linux-64<2023.0.0'
in the conda install commmand.The text was updated successfully, but these errors were encountered: