Repository navigation
Conversation
Other programs on the card take from everything a run measures. On the reference RTX 4070 an editor that stood open, drawing 2.2 % of the 3D engine while idle, cost 8 % of the stress score and 6 % of the memory bandwidth for a whole day of tests; a video playing in a browser made the bandwidth readings jump by 3 to 8 %. Before a run the tool now reads the card's use per program from Windows' own counters for three seconds. Programs above 1 % are named. The window asks what to do: check again, go on anyway, or cancel. The command line names them and does not start; --busy-ok as the last argument lets it. Nothing is ever closed by the tool, and such a program is no reason to refuse a run, as another tuning tool is: it does nothing to the card. Every look is written to the run's log, so the log says what was on the card. After an optimize run a second look says whether that changed under it; the result then carries a note that its comparison with stock is not to be trusted. gao --busy shows the programs a run would find, and changes nothing. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The elevated check of hardware check 96 (d) passed: next to an open editor the run named it, did not start, and changed nothing. It also showed the editor reading 1 % in one look and 2 % in the next, on the line that decided whether it was named at all. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Oct 11, 2026
Merged
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
1 participant
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.
What
Before a run the tool reads the card's use per program from Windows' own counters for three seconds and names every program at 0.5 % or more.
--busy-okas the last argument lets it.gao --busyshows what a run would find and changes nothing.Why
For a whole day of card tests the reference RTX 4070 scored about 5300 it/s and read 444 to 455 GB/s of memory bandwidth, scattering, until an editor that had stood open since the morning was closed: 5830 it/s and 480 GB/s, steady, from that second on. The editor was idle and used 2.2 % of the 3D engine. A video playing in a browser makes the bandwidth readings jump by 3 to 8 %. Nobody thinks of an open editor as a heavy program, so the tool has to look.
Tested
tests/test_busy_card.cpp) and for the counter-name parser (tests/test_gpu_users.cpp); suite and AddressSanitizer suite green.gao --busynames the editor; elevated,gao --retestnext to it names it, exits 1 and changes nothing.--busy-ok, and the three answers in the window (seen in the demo mode of the next PR).🤖 Generated with Claude Code