Update graph_optimization.ipynb for Intel Xeon 6 Processors

[louie-tsai]

Intel Xeon 6 Processors also support fp16 with auto mixed precision optimizer

[github-actions]

Preview

Preview and run these notebook edits with Google Colab:

• site/en/guide/graph_optimization.ipynb

Rendered notebook diffs available on ReviewNB.com.

Format and style

Use the TensorFlow docs notebook tools to format for consistent source diffs and lint for style:

$ python3 -m pip install -U --user git+https://github.com/tensorflow/docs

$ python3 -m tensorflow_docs.tools.nbfmt notebook.ipynb

$ python3 -m tensorflow_docs.tools.nblint --arg=repo:tensorflow/docs notebook.ipynb

If commits are added to the pull request, synchronize your local branch: git pull origin graph_optimze

[8bitmp3]

@MarkDaoust @markmcd

[MarkDaoust]

Hi @louie-tsai, I don't know anything about this. Do you have any references showing that this is true?

[louie-tsai]

Hi @louie-tsai, I don't know anything about this. Do you have any references showing that this is true?

Thanks for asking.
Here are couple references for fp16 on the latest Xeon processors.
https://www.intel.com/content/www/us/en/content-details/669773/intel-avx-512-fp16-instruction-set-for-intel-xeon-processor-based-products-technology-guide.html
https://www.intel.com/content/www/us/en/newsroom/news/intel-unveils-future-generation-xeon.html#gs.g4yc8z
https://oneapi-src.github.io/oneDNN/enum_dnnl_cpu_isa.html

[MarkDaoust]

Cool, thanks. I see the processors have FP16. Does anyone know for sure that this actualy works with the Auto mixed precision optimizer?

[louie-tsai]

Cool, thanks. I see the processors have FP16. Does anyone know for sure that this actualy works with the Auto mixed precision optimizer?

It works for AMP optimizer.
Loop the Intel architect @agramesh1 for more comment if any.

[agramesh1]

Cool, thanks. I see the processors have FP16. Does anyone know for sure that this actualy works with the Auto mixed precision optimizer?

It works for AMP optimizer. Loop the Intel architect @agramesh1 for more comment if any.

Hi @MarkDaoust yes, it is now supported in TF and XLA. We have been working with @penpornk to add support for fp16 AMP using the AMX-FP16 that was introduced in the latest Intel Xeon CPUs.

[MarkDaoust]

Cool thanks!

[louie-tsai]

@markmcd
possible to help review and approve the PR?
thanks!

[markmcd]

@markmcd possible to help review and approve the PR? thanks!

Done - copybara will merge shortly.