Update bfloat16 support for Intel CPUs

[gaurides]

This PR updates mixed_precision documentation to reflect that Intel CPU Sapphire Rapids supports BFloat16 optimizations using AMX instructions and user can expect performance boost.
It doesn't affect any other part of the document.

[github-actions]

Preview

Preview and run these notebook edits with Google Colab:

• site/en/guide/mixed_precision.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 gaurides/mixedprecision_doc

[gaurides]

@penpornk : Can you please review this PR to update documentation?

[gaurides]

@penpornk, @8bitmp3 - Please let me know if you have any questions. Thanks.

[penpornk]

@MarkDaoust Could you please help review this mixed precision doc update? :)
Contents look okay to me. Wording-wise, I'm not sure if some parts are clear enough to general TF readers (or if they need more explanations).

[MarkDaoust]

Thanks for the ping. I think its okay, let's get this merged, it's been waiting long enough.

[gaurides]

@MarkDaoust - I included your suggestion. Can you please re-approve? Thanks.

[MarkDaoust]

Thanks.