How to configure onAPI to use NVIDIA GPU driver?

[Caster]

Although I was able successfully compile and run example Base_Vector_Add (SYCL) on my notebook, I got an error (output, not a code):

Running on device: Intel(R) HD Graphics 620
Vector size: 10000
LLVM ERROR: SPIRV internal error.

How to configure onAPI to use NVIDIA GPU driver and not Intel GPU Driver?

[ArpanB_Intel]

Hi Dan, could you confirm the full name of the Intel® oneAPI Toolkit you are using and it's version with us?

We would like to know.

[Caster]

On my notebook (Windows 10, 64 bits) with the latest version of VisualStudio 2022 and NVIDA drivers (CUDA 11.8 Runtime - VS22 template), I installed the one API Base Toolkit which is now part of the VS22.

I am writing and running CUDA programs in VS22 (CUDA 11.8 Runtime - VS22 template) for my NVIDIA GeForce GTX 960M without any problems.

"About" returns Intel VTune Profiler 2023.0.0, Product build: 624757

[ArpanB_Intel]

Dan, we suggest you to click here: https://www.intel.com/content/www/us/en/developer/articles/system-requirements/vtune-profiler-system-requirements.html#inpage-nav-3, and check the Software Requirements section and the Graphics Driver Installation section for further information.

[Caster]

Thank you. There is mentioned:

Graphics Driver Installation:

Intel® oneAPI 2023.0 is validated on Windows and Linux for the following driver versions

• Windows Intel® Graphics Driver
  To install the driver follow the directions in the article appropriate for your device
  • Intel® Iris® Xe MAX (DG1) and 10th-13th Gen Intel® Core™ Processor Graphics https://www.intel.com/content/www/us/en/download/19344/intel-graphics-windows-dch-drivers.html
  • Intel® Arc™ A-Series Graphics (DG2) https://www.intel.com/content/www/us/en/download/726609/intel-arc-graphics-windows-dch-driver.html
  • Intel® Data Center GPU Flex Series (ATS-M) https://registrationcenter.intel.com/en/products/subscription/956/
  • Contact your OEM representative for access to the Intel Registration Center.

Does it mean I can't use one API for generating SYCL program and run it on my NVIDIA GPU?

I have wsl (Debian) subsystem on my notebook too and can run SYCL program on Debian system either (boot the external SSD drive attached to my notebook).

[ArpanB_Intel]

Dan, generally our products are tested on a selected list of platforms prior to their release. This is the list of supported platforms for the toolkit.

The toolkit may or may not work on other platforms. However, if it does not work, we will be unable to support them.

[ArpanB_Intel]

Dan, unfortunately we were unable to hear back from you.

If you have any further queries, please post a new question as this thread will no longer be monitored by Intel®.