ci: Update Isaac Sim to latest digest (from 08/06/2026) - #6966
Merged
Conversation
Kit 9e5dd32c (OMPE-65433) flips /rtx/dldenoiser/responsiveDenoising to true, which shifts every RTX-rendered golden. The change is intentional upstream work to reduce ghosting, so take the new build and regenerate the goldens instead of holding CI on the pre-change pin.
The new DLSS-RR default shifts every RT2-rendered frame, so the rgb and rgba AOVs land outside threshold on all six rendering-correctness scenes and on the Kit visualizer captures. Every other AOV stays under. Goldens are taken from the rendering-correctness and isaaclab_visualizers jobs of the run against the newly pinned image.
Kit 9e5dd32c keeps changing DLSS-RR pixels while simulation is paused. Disable responsive denoising only in this integration-test process so pause behavior remains testable without changing application defaults.
Responsive DLSS-RR continues temporal accumulation on static frames, so pixel identity is not a valid pause signal for Kit on Newton. Validate the simulation flag, physics-step counter, and body state instead while keeping strict frame checks for deterministic render paths.
Track the responsive-denoising regression with NVBug 6570125 instead of silently bypassing its assertion. Split the visualizer cases so Newton's tiled integration path remains fully covered.
Test Isaac Sim commit 307e5486 by exact digest so repeated CI runs use the same image.
Use the stable outputs captured by the latest CI run after the visualizer and locomotion updates on develop.
Kit c0b875cd breaks opt-in per-environment scene partitions. Keep the registered rendering checks on the default path and xfail the dedicated partition tests until Kit is fixed.
Use the PhysX outputs captured by the follow-up CI run.
nvsekkin
commented
Aug 7, 2026
nvsekkin
commented
Aug 7, 2026
Use the RGB and RGBA outputs captured by the current latest-image run.
The Kit case now passes on both digest branches and in the latest-image local reproduction.
nvsekkin
changed the title
nvsekkin
marked this pull request as ready for review
August 8, 2026 00:32
nvsekkin
requested review from
hujc7,
kellyguo11,
myurasov-nv and
ooctipus
as code owners
August 8, 2026 00:32
nvsekkin
requested review from
AntoineRichard,
StafaH and
matthewtrepte
as code owners
August 8, 2026 00:32
Contributor
Greptile SummaryThe PR pins CI to a newer Isaac Sim digest, refreshes renderer golden images, and temporarily quarantines the known scene-partition regression.
Confidence Score: 4/5The PR appears safe to merge, with the non-blocking caveat that its non-running xfails cannot signal when the quarantined regression is fixed. The image and golden updates have no established blocking defect, while the partition quarantine should execute its tests to preserve XPASS and changed-failure visibility. Files Needing Attention: source/isaaclab_physx/test/renderers/test_isaac_rtx_renderer_scene_partitioning.py Important Files Changed
Reviews (1): Last reviewed commit: "test: Document skip fragment reasons" | Re-trigger Greptile |
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
Comment on lines
+53
to
+55
| _PARTITION_XFAIL = pytest.mark.xfail( | ||
| reason="NVBug 6264822: Kit c0b875cd scene partitioning regression", strict=False, run=False | ||
| ) |
Contributor
There was a problem hiding this comment.
Setting run=False prevents both partition-isolation tests from executing, so CI cannot produce an XPASS when NVBug 6264822 is fixed or expose a changed failure mode. Keep the non-strict xfail while allowing the test bodies to run.
Suggested change
| _PARTITION_XFAIL = pytest.mark.xfail( | |
| reason="NVBug 6264822: Kit c0b875cd scene partitioning regression", strict=False, run=False | |
| ) | |
| _PARTITION_XFAIL = pytest.mark.xfail( | |
| reason="NVBug 6264822: Kit c0b875cd scene partitioning regression", strict=False | |
| ) |
Note: If this suggestion doesn't match your team's coding style, reply to this and let me know. I'll remember it for next time!
![isaaclab-review-bot[bot]](https://avatars.githubusercontent.com/u/157846462?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Isaac Lab Review Bot
Reviewed the Isaac Sim digest repin, refreshed rendering goldens, and the targeted quarantine of the two scene-partitioning tests. The registered-task rendering suite continues on the default non-partitioned path while the known Kit regression is tracked separately.
- Design and architecture: The quarantine is limited to the two tests that require scene partitioning. The unchanged default-off test still verifies the feature gate, and registered-task rendering continues to cover the default renderer path. The primary tradeoff is temporarily losing executable scene-partition isolation coverage until NVBug 6264822 is resolved.
- API: No public APIs, configuration schemas, CLI arguments, or exports change. The test-only package changes include the required
.skipchangelog fragments, and the CI image configuration remains digest-pinned for reproducibility. - Implementation: The shared
xfailmarker usesrun=False, preventing the affected test bodies and fixtures from executing, while fixture removal from the registered-task test deliberately restores default rendering behavior. Updated golden pointers align with the new Isaac Sim image and selected rendering path. Because the quarantine is non-strict and does not run, CI cannot detect recovery automatically; restoration depends on the linked NVBug and TODO being followed up.
No blocking issues. No inline issue met the actionable-evidence threshold; the assessment above records the review feedback.
Automated review; human maintainers own approval decisions.
kellyguo11
approved these changes
Aug 8, 2026
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.
Summary
307e5486at digestsha256:a13867a6403488a9e640288f96492e6ecd44c3a525b5b85793a2da68fdac48ad.c0b875cdscene-partition regression tracked by NVBug 6264822 while keeping registered rendering coverage on the default path.Type of change
Checklist
pre-commitchecks with./isaaclab.sh --formatsource/<pkg>/changelog.d/for every touched package (do not editCHANGELOG.rstor bumpextension.toml— CI handles that)CONTRIBUTORS.mdor my name already exists there