Fix incorrect channel ordering in FusedSSIM prefetch - #750
Merged
Conversation
Signed-off-by: Matthew Cong <mcong@nvidia.com>
Signed-off-by: Matthew Cong <mcong@nvidia.com>
Signed-off-by: Matthew Cong <mcong@nvidia.com>
matthewdcong
force-pushed
the
fix_ssim_prefetch_order
branch
from
August 21, 2026 22:51
111b939 to
a58bc45
Compare
swahtz
reviewed
Aug 22, 2026
swahtz
approved these changes
Aug 22, 2026
swahtz
left a comment
swahtz
left a comment
Contributor
There was a problem hiding this comment.
Looks great. Small nit that a Codex review found but otherwise lgtm
Signed-off-by: Matthew Cong <mcong@nvidia.com>
matthewdcong
force-pushed
the
fix_ssim_prefetch_order
branch
from
August 22, 2026 01:51
50b8121 to
92b944e
Compare
matthewdcong
enabled auto-merge (squash)
August 22, 2026 01:52
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Previously, prefetching in FusedSSIM assumed a NHWC ordering which matches the storage ordering of the image on disk. However, the image is permuted during the training pipeline to NCHW in order to leverage the fused SSIM kernels. This incorrect prefetching led to a significantly amount of unnecessary UVM traffic and fixing it improves fused SSIM scalability dramatically. Measured on 2x RTX 3090s: