Skip to content

Optimize Xtensa transpose_conv and decompress kernels - #3727

Open
unmeshna017 wants to merge 1 commit into
tensorflow:mainfrom
unmeshna017:xtensa-transpose-conv-decompress
Open

unmeshna017 wants to merge 1 commit into
tensorflow:mainfrom
unmeshna017:xtensa-transpose-conv-decompress

Conversation

@unmeshna017

Copy link
Copy Markdown
Contributor

Adds NNLib-optimized transpose_conv (float32, int8 and int16) and minor decompress/decompress_common cleanups.

Depends on the NNLib 09_03_2026 release (PR#3703) for the v2 transpose_conv APIs.

@cad-audio @joshih-cad

Adds HiFi5 NNLib-optimized transpose_conv (float32, int8 and int16) using the xa_nn_transpose_conv_v2_sym8sxasym8s/_sym16s APIs, and minor decompress/decompress_common cleanups. Depends on the NNLib 09_03_2026 release (PR#3703) for the v2 transpose_conv APIs.
@unmeshna017
unmeshna017 requested a review from a team as a code owner September 17, 2026 07:14

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant