[PATCH] D121510: [Support] Introduce the BLAKE3 hashing function implementation

Tue Mar 15 15:44:26 PDT 2022

akyrtzi added a comment.

In D121510#3381583 <https://reviews.llvm.org/D121510#3381583>, @MaskRay wrote:

> It will be worth mentioning which optimized implementation is used for Intel Xeon W and M1Pro, respectively.
>
> The performance of using the generic implementation is also interesting. None of LLVM's SHA1/SHA256/MD5 libraries is well optimized.

Due to an issue with detecting AVX512 on macOS <https://github.com/BLAKE3-team/BLAKE3/issues/232> the numbers for Intel I posted initially were for `AVX2`. Here are new speed-up multiplies that include `AVX512` and the generic/portable implementations:

| Processor               | SHA1  | SHA256 | MD5  |
| ----------------------- | ----- | ------ | ---- |
| Intel Xeon W (AVX512)   | 10.4x | 27x    | 9.4x |
| Intel Xeon W (AVX2)     | 6.5x  | 17x    | 5.9x |
| Intel Xeon W (portable) | 1.3x  | 3.3x   | 1.1x |
| M1Pro (neon)            | 2.1x  | 4.7x   | 2.8x |
| M1Pro (portable)        | 1.1x  | 2.4x   | 1.5x |
|

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D121510/new/

https://reviews.llvm.org/D121510