- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Reference:
Solved: Re:Deep Neural Network Training on GPU - Intel Community
I want to train neural network in Intel Iris Plus iGP.
The introduction video on oneAPI showed intel optimizations of pytorch and tensorflow uses oneAPI (oneDNN).
Since my computer only supports the Intel iris Plus iGP, will intel-pytorch and intel-tensorflow enable me to do the training work?
Also, how are the performance comparisons between nvidia and intel iris plus on (training) neural network.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Aswin,
Yes, you could use Intel Xeon scalable processors in devcloud for training. Devcloud also has Intel optimized versions of tensorflow and pytorch preinstalled.
You could run your training workloads on below hardware(CPU) options.
Intel® Xeon® Scalable 6128 processors
Intel® Xeon® Scalable 8256 processors
Intel® Xeon® E-2176 P630 processors (with Intel® Graphics Technology)
Thanks
Arun
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Aswin,
Intel optimised frameworks tensorflow and pytorch currently supports only CPU. Thus you would not be able to run your training workloads on Intel Iris Plus or any other Intel GPUs currently. Please find the system requirements for AI analytics toolkit which mentions support for Intel® Graphics (GPU) to be added in future release.
For using OneDNN directly the system requirements mention to have GPUs which are gen9 and above, or Xe architecture.
https://software.intel.com/content/www/us/en/develop/tools/oneapi/components/onednn.html#gs.d33fft
Thanks
Arun
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Can the Intel Dev cloud be used to train neural network models using Intel PyTorch and Intel Tensorflow. A valuable resource says Intel Xeon processors are available for training.
Could this be a better alternative?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Aswin,
Yes, you could use Intel Xeon scalable processors in devcloud for training. Devcloud also has Intel optimized versions of tensorflow and pytorch preinstalled.
You could run your training workloads on below hardware(CPU) options.
Intel® Xeon® Scalable 6128 processors
Intel® Xeon® Scalable 8256 processors
Intel® Xeon® E-2176 P630 processors (with Intel® Graphics Technology)
Thanks
Arun
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Aswin,
Hope we have answered your queries, please let us know if you need any other help regarding this.
Thanks
Arun
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Thanks, that was really helpful in making a decision. I got signed up to Intel DevCloud for further training and accessing HPC resources.
The DevCloud was of a great help even during the completion of A Nanodegree Scholarship with Intel and Udacity.
My university also offers HPC cluster resource which is after I try out things in Intel DevCloud.
That's a great resource from Intel to the Public.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Aswin,
We are glad the solution provided helped and thank you for your feedback on devcloud. As the current issue If you need any additional information, please submit a new question as this thread will no longer be monitored
Thanks
Arun
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page