---
title: "Discussion on \"PyTorch 2.13 on Apple Silicon: FlexAttention is up to 12x faster than SDPA\" | Hashnode"
description: "Discussion on \"PyTorch 2.13 on Apple Silicon: FlexAttention is up to 12x faster than SDPA\". PyTorch 2.13 on Apple Silicon: FlexAttention is up to 12x faster than SDPA\nSummary. PyTorch 2.13, released around July 11, 2026, brings FlexAttention to Apple Silicon through the Metal (MPS) backend,"
url: "https://hashnode.com/posts/pytorch-213-on-apple-silicon-flexattention-is-up-to-12x-faster-than-sdpa/6a5baeef299d84cdace2a7dd"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [PyTorch 2.13 on Apple Silicon: FlexAttention is up to 12x faster than SDPA](https://ecorpit.hashnode.dev/pytorch-213-on-apple-silicon-flexattention-is-up-to-12x-faster-than-sdpa?utm_source=hashnode&utm_medium=feed)

PyTorch 2.13 on Apple Silicon: FlexAttention is up to 12x faster than SDPA Summary. PyTorch 2.13, released around July 11, 2026, brings FlexAttention to Apple Silicon through the Metal (MPS) backend,

Jul 18 · 13 min read · ecorpit.hashnode.dev  
Read the article

-   [#pytorch](https://hashnode.com/tag/pytorch)
-   [#apple-silicon](https://hashnode.com/tag/apple-silicon)
-   [#flexattention](https://hashnode.com/tag/flexattention)
-   [#mps](https://hashnode.com/tag/mps)

## Written by

[Manu Shukla](https://hashnode.com/@mr-manushukla)

@mr-manushukla · Delhi, India

  
[View profile](https://hashnode.com/@mr-manushukla)

Founder and Director of eCorpIT

## Responses

Markdown supported · ⌘↵ to post

No responses yet.