Benchmarks found: 29
Seeing how coalescing impacts performance metrics.
By ritzdacatTesting whether cpu reductions on the scheduler side impact framerate.
By ritzdacatReviewing to see if reducing the cpu usage of the scheduler for scx_gamer increased framerate or improved metrics compared to previous instances. Included previous scheduler data for comparison. 4k resolution, firing range, aiming at 40 target, shadows low, reflections low, rest of settings on high. TSR ultra performance.
By ritzdacatTesting some optimization where llvm-objdump is used to make some refactors to the scheduler performance.
By ritzdacatA: LTO FAT CPU Native B: lto fat + cpu native + ringbuffer changes
By ritzdacatSplitgate 2 - Testing LTO:FAT and CPU:Native LTO:Fat and CPU:Native + OC 200 core and 2000 mem was on 11/15 rest are older captures to compare.
By ritzdacat4k firing range targeting same sign 3 runs of each scheduler. Beerland needs a patch in 6.18 and my kernel is still 6.17
By ritzdacatsame test as before but with scx_gamer patch b to test some changes
By ritzdacatfiring range arc raiders 4k resolution, high settings. Testing various schedulers and versions then testing again with core clock +200 and memclock + 2000 to see how it impacts Arc Raiders. SCX Gamer has a modification that should allow us to force dispatch to other cpus and remove afinitys that the RHI has set, will see if that helps or made things worse.
By ritzdacat