Add AArch64-optimized SYMV kernels #5221
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This pull request adds [SD]SYMV kernels optimized for arm64.
Previously, generic/symv_k.c has been used on arm64 systems, in which the calculation falls back to using the GEMV kernels. This approach involves accessing each matrix element twice. To address this inefficiency, I’ve implemented new kernels by using a technique analogous to those employed in the x86_64 SYMV kernels.
As shown in the graphs below, performance is improved by about 2x on the A64FX, Graviton3E, Grace, and Ampere Altra Max platforms, respectively.
SYMV is an important component in symmetric matrix eigenvalue computations. Consequently, this PR yields performance improvements in higher-level routines such as DSYEVD. Specifically, the execution time of DSYTRD (tridiagonalization), which is the initial step within DSYEVD, is reduced as shown in the graph.