wyj
7ddc58515a
Document the producer-private adaptive path, the dummy support workload diagnostics, the negative result of the old serial binning and the 2026-09-13 full-day comparison (splat 282.944 s vs atomic 481.417 s). Add the corresponding raw logs, hotspot fixtures and runner evidence, and lay out the staged plan toward a sub-240 s full-day render without changing the default atomic path.
10 lines
1.4 KiB
Plaintext
10 lines
1.4 KiB
Plaintext
COMMAND OMP_NUM_THREADS=16 OMP_DYNAMIC=FALSE env HIP_PSF_ACCUMULATION=atomic build/Release/replay_psf benchmarks/hip_psf_replay_2026-09-07/center_hotspot_jitter8_65536.events 65536 production 128
|
|
TIMEOUT 45s
|
|
GIT 2d01c2c101ccea8193a64bb46ceafca3009f0ef5
|
|
BINARY_SHA256 d056d76de917e764c4f1d516a51657c8d9e9cf3adf97615ccb50b6c81a565002
|
|
PSF cache ready: 64x64 phases, radius 47 px, relative tail 1e-08, tail abs 1e-06, boundary 1e-07, build 0.700 s
|
|
device=AMD Radeon RX 9070 arch=gfx1201 events=65536 rounds=128 frame=3840x2160 mode=production distribution=captured contributions_per_round=432818979 CPU_reference=0.825222015 create=0.039162159
|
|
HIP PSF progress: 6455296 submitted / 6422528 completed events; 392 / 394 batches timed; kernel 4.950 s
|
|
complete_replay=6.520941973 replay_wall=6.481779814 select=0.000000000 bin=0.000000000 upload=0.025792579 prepare=0.000000000 accumulation=6.446903917 merge=0.000000000 download=0.004890707 cleanup=0.000000000 events_s=1294182.808 contributions_s=8547163110.098 center_tiles=0 tile_chunks=0 atomic_chunks=512 refs=0 tasks=0 scratch_used_peak=0 scratch_device_reserved=0 RSS_KiB=706148 max_abs=2.0834028946481453e-14 max_rel=1.8520792977078497e-12 flux_R=-7.2487421955847032e-14 flux_G=9.3608940832863565e-14 flux_B=6.0115331056700439e-14 flux_Y=3.2639931870553769e-14 validation=scaled-cpu-double-reference
|
|
PROCESS_WALL 8.378555340 EXIT 0
|