Benchmarks found: 54
main branch, 0ad2f30 commit (latest at the time of writting). Background load: stress -c 8 (steam deck has 8 cores (probably 4c8t)
By erikas2main branch, 0ad2f30 commit (latest at the time of writting)
By erikas2in-game camera spin test
By erikas2in-game camera spin test
By erikas2in-game camera spin test
By erikas2Using 6.10.2-2-cachyos-deckify-lto. "bpfland_args"is "scx_bpfland -t 0 -c 0", "bpfland_pd" is "scx_bpfland --preemption-disabled", "lavd_ncc" is "scx_lavd --no-core-compaction". Used in-game camera spin for 30 seconds, as benchmark does not cause CPU stutter (probably used to benchmark graphics performance only).
By erikas2bpfland_args is bpfland with "-t 0 -c 0" args. Tested on Steam Deck. Basically use joystick to spin the camera around non-stop. Simulates in-game lag, which is not visible in integrated benchmark.
By erikas2bpfland_args is bpfland with "-t 0 -c 0" args. Tested on Steam Deck.
By erikas2I am uploading this benchmark to understand FPS stability of different schedulers. I state that bpfland gives SUPERIOR fps stability to the point that there are no stutters at all. Remember when you set FPS limit to 40 or 30fps, so it would be stable? That's right, except bpfland gives such stable fps without any need for FPS cap. Let's see if numbers would prove me right. :D
By erikas2Basically 'echo "manual" > /sys/class/drm/card0/device/power_dpm_force_performance_level' and each option in /sys/class/drm/card0/device/pp_power_profile_mode
By erikas2