- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
I am installing OpenVino 2020.2 from apt for my Ubuntu 18.04 system but there is no setup for my GPU driver available?
Hoewer would it make sense to run it on the integrated Graphics Controller (Celeron N3350/Pentium N4200/Atom E3900 Series) in my Up Squared Ai Edge?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks for reaching out.
Intel® Graphics Compute Runtime for OpenCL™ is not a part of OpenVINO™ APT distribution. You can install it from the Intel® Graphics Compute Runtime for OpenCL™ GitHub repo.
Let us know in case of any other questions.
Regards,
Ram prasad
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks for reaching out.
Intel® Graphics Compute Runtime for OpenCL™ is not a part of OpenVINO™ APT distribution. You can install it from the Intel® Graphics Compute Runtime for OpenCL™ GitHub repo.
Let us know in case of any other questions.
Regards,
Ram prasad
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks, that worked!
However, when the model is loaded into the device (took quiet some time) and inference is started my GPU is making weird sounds. Is the model somehow too heavy then?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
You can use cl_cache feature to reduce the loading time for GPU. However, for the first time the loading time will be same and then faster. Refer FAQ(OpenCL).
What is the size of the model?
Are you using OpenVINO pre-trained model?
Regards,
Ram prasad
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I was trying the pretrained MobileNetV2 from person-detection-retail-0013. And I ran it using the pedestrian detection demo.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
We have tried similar pre-trained models on the same setup but haven't observed the mentioned issue. This seems like a hardware specific issue. Please reach out to the corresponding vendor to address this issue better.
Regards,
Ram prasad

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page