Skip to content

perf: compile SwiftVR ReAE compute blocks - #629

Draft
gtong-nv wants to merge 2 commits into
mainfrom
dev/gtong/swiftvr
Draft

gtong-nv wants to merge 2 commits into
mainfrom
dev/gtong/swiftvr

Conversation

@gtong-nv

@gtong-nv gtong-nv commented Sep 15, 2026

Copy link
Copy Markdown
Collaborator

Configuration and performance chart

Candidate Transformer Encoder Decoder Median p90 Effective FPS Cold prepare Peak allocation
Eager off off off 96.47 ms 96.86 ms 82.9 3.2 s 18.65 GiB
Encoder off on off 92.31 ms 92.66 ms 86.7 24.7 s 18.65 GiB
Decoder off off on 82.96 ms 83.26 ms 96.4 121.2 s 20.45 GiB
Encoder + decoder off on on 79.91 ms 80.57 ms 100.1 145.5 s 20.45 GiB
Transformer on off off 94.69 ms 95.34 ms 84.5 30.0 s 18.65 GiB
Transformer + decoder on off on 79.60 ms 79.97 ms 100.5 142.0 s 20.45 GiB

Signed-off-by: Gangzheng Tong <gtong@nvidia.com>
@copy-pr-bot

copy-pr-bot Bot commented Sep 15, 2026

Copy link
Copy Markdown

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

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