rustytorch

embedded3 PERF-3 P1-5 (c) part 2 + (d): the device tables persist across steps and are updated in place within the band — one DeviceCut with both phases' aperture / open / activity buffers (the predictor phase is a flag); the faces of the changed cell…

Documentation #1011:Commit 0efb8be745 pushed by osobh

main

2026-09-20 23:17:39 +00:00

6s

embedded3 PERF-3 P1-5 (c) part 2 + (d): the device tables persist across steps and are updated in place within the band — one DeviceCut with both phases' aperture / open / activity buffers (the predictor phase is a flag); the faces of the changed cell…

CI #1010:Commit 0efb8be745 pushed by osobh

main

2026-09-22 02:32:42 +00:00

27h15m11s

embedded3 PERF-3 P1-5 (c) part 2 + (d): the device tables persist across steps and are updated in place within the band — one DeviceCut with both phases' aperture / open / activity buffers (the predictor phase is a flag); the faces of the changed cell…

Performance Benchmarks #1009:Commit 0efb8be745 pushed by osobh

main

2026-09-20 23:20:37 +00:00

3m6s

embedded3 PERF-3 P1-5 (c) part 1: scatter kernels (e3_scat_{f64,f32,u32,i32}) in the CG module; the refresh's seven f64 coefficient arrays and z links, and the V-cycle fine level's f32 arrays and maps, scattered on the changed rows into persistent buff…

Documentation #1008:Commit 77fda442c6 pushed by osobh

main

2026-09-20 23:09:41 +00:00

31s

embedded3 PERF-3 P1-5 (c) part 1: scatter kernels (e3_scat_{f64,f32,u32,i32}) in the CG module; the refresh's seven f64 coefficient arrays and z links, and the V-cycle fine level's f32 arrays and maps, scattered on the changed rows into persistent buff…

CI #1007:Commit 77fda442c6 pushed by osobh

main

2026-09-22 02:32:42 +00:00

27h23m33s

embedded3 PERF-3 P1-5 (c) part 1: scatter kernels (e3_scat_{f64,f32,u32,i32}) in the CG module; the refresh's seven f64 coefficient arrays and z links, and the V-cycle fine level's f32 arrays and maps, scattered on the changed rows into persistent buff…

Performance Benchmarks #1006:Commit 77fda442c6 pushed by osobh

main

2026-09-20 23:12:37 +00:00

3m28s

embedded3 PERF-3 P1-5 (a)(b): the previous-step aperture/volume records move out of the old mask instead of being cloned; the GCL table over the changed cells and the wall cells (the serial sums with their exact-zero terms left out, same order) — slab…

Documentation #1005:Commit e3958fc9c5 pushed by osobh

main

2026-09-20 22:56:45 +00:00

57s

embedded3 PERF-3 P1-5 (a)(b): the previous-step aperture/volume records move out of the old mask instead of being cloned; the GCL table over the changed cells and the wall cells (the serial sums with their exact-zero terms left out, same order) — slab…

CI #1004:Commit e3958fc9c5 pushed by osobh

main

2026-09-22 02:32:42 +00:00

27h36m57s

embedded3 PERF-3 P1-5 (a)(b): the previous-step aperture/volume records move out of the old mask instead of being cloned; the GCL table over the changed cells and the wall cells (the serial sums with their exact-zero terms left out, same order) — slab…

Performance Benchmarks #1003:Commit e3958fc9c5 pushed by osobh

main

2026-09-20 22:55:50 +00:00

6s

embedded3 PERF-3 P1-4 step e: Components::summary_planes — the component count and singular flags alone on the device path (no per-cell id / member arrays); slab CSV byte-identical, device moving/cg green; ny 124 rebuild block 2,119 → 1,995 ms per st…

Documentation #1002:Commit c6a35a77e2 pushed by osobh

main

2026-09-20 18:54:53 +00:00

12s

embedded3 PERF-3 P1-4 step e: Components::summary_planes — the component count and singular flags alone on the device path (no per-cell id / member arrays); slab CSV byte-identical, device moving/cg green; ny 124 rebuild block 2,119 → 1,995 ms per st…

CI #1001:Commit c6a35a77e2 pushed by osobh

main

2026-09-21 20:32:42 +00:00

25h38m2s

embedded3 PERF-3 P1-4 step e: Components::summary_planes — the component count and singular flags alone on the device path (no per-cell id / member arrays); slab CSV byte-identical, device moving/cg green; ny 124 rebuild block 2,119 → 1,995 ms per st…

Performance Benchmarks #1000:Commit c6a35a77e2 pushed by osobh

main

2026-09-20 18:57:35 +00:00

2m55s

embedded3 PERF-3 P1-4 (steps 1–3 + d): the band-persistent operator — CutGeometry marks re-evaluated corners; Mask::changed_cells (touched by either build, dilated by one); Level::patch re-derives the changed rows with the constructor's formulas on t…

Documentation #999:Commit fe76413c66 pushed by osobh

main

2026-09-20 18:47:50 +00:00

1m10s

embedded3 PERF-3 P1-4 (steps 1–3 + d): the band-persistent operator — CutGeometry marks re-evaluated corners; Mask::changed_cells (touched by either build, dilated by one); Level::patch re-derives the changed rows with the constructor's formulas on t…

CI #998:Commit fe76413c66 pushed by osobh

main

2026-09-21 20:32:42 +00:00

25h46m4s

embedded3 PERF-3 P1-4 (steps 1–3 + d): the band-persistent operator — CutGeometry marks re-evaluated corners; Mask::changed_cells (touched by either build, dilated by one); Level::patch re-derives the changed rows with the constructor's formulas on t…

Performance Benchmarks #997:Commit fe76413c66 pushed by osobh

main

2026-09-20 18:46:44 +00:00

7s

embedded3 PERF-3 P1-3 (f): the mask build's whole-grid scans as parallel per-entry maps (cell classification with the anchor as the smallest fluid index, the three face-kind tables, the merging scan's small flags, the centroid-shift table per face) — …

Documentation #996:Commit 44caeb8110 pushed by osobh

main

2026-09-20 17:31:11 +00:00

6s

embedded3 PERF-3 P1-3 (f): the mask build's whole-grid scans as parallel per-entry maps (cell classification with the anchor as the smallest fluid index, the three face-kind tables, the merging scan's small flags, the centroid-shift table per face) — …

CI #995:Commit 44caeb8110 pushed by osobh

main

2026-09-21 20:32:42 +00:00

27h1m38s

embedded3 PERF-3 P1-3 (f): the mask build's whole-grid scans as parallel per-entry maps (cell classification with the anchor as the smallest fluid index, the three face-kind tables, the merging scan's small flags, the centroid-shift table per face) — …

Performance Benchmarks #994:Commit 44caeb8110 pushed by osobh

main

2026-09-20 17:33:08 +00:00

2m4s

embedded3 PERF-3 P1-3 (c)(d)(e): the moving path's host mirror persists between steps and only the arrays the rebuild reads come down (u, v, w, p, the previous velocities) and only those it writes go up (u, v, w, p — the correction stays on the device…

Documentation #993:Commit 5a376630a7 pushed by osobh

main

2026-09-20 17:24:08 +00:00

58s

embedded3 PERF-3 P1-3 (c)(d)(e): the moving path's host mirror persists between steps and only the arrays the rebuild reads come down (u, v, w, p, the previous velocities) and only those it writes go up (u, v, w, p — the correction stays on the device…

CI #992:Commit 5a376630a7 pushed by osobh

main

2026-09-21 20:32:42 +00:00

27h9m36s

embedded3 PERF-3 P1-3 (c)(d)(e): the moving path's host mirror persists between steps and only the arrays the rebuild reads come down (u, v, w, p, the previous velocities) and only those it writes go up (u, v, w, p — the correction stays on the device…

Performance Benchmarks #991:Commit 5a376630a7 pushed by osobh

main

2026-09-20 17:23:14 +00:00

8s

embedded3 PERF-3 P1-3 (a): the operator refresh moves the operator into the level (no eight-array clone) and takes its components/links from the level's problem; Components::find_parallel (hook-and-shortcut union-find, same numbering) on the refresh an…

Documentation #990:Commit 651f75d96a pushed by osobh

main

2026-09-20 16:57:32 +00:00

14s

embedded3 PERF-3 P1-3 (a): the operator refresh moves the operator into the level (no eight-array clone) and takes its components/links from the level's problem; Components::find_parallel (hook-and-shortcut union-find, same numbering) on the refresh an…

CI #989:Commit 651f75d96a pushed by osobh

main

2026-09-21 20:32:42 +00:00

27h35m25s

embedded3 PERF-3 P1-3 (a): the operator refresh moves the operator into the level (no eight-array clone) and takes its components/links from the level's problem; Components::find_parallel (hook-and-shortcut union-find, same numbering) on the refresh an…

Performance Benchmarks #988:Commit 651f75d96a pushed by osobh

main

2026-09-20 16:59:30 +00:00

2m14s

embedded3 S2-7b: the moving-wall forms on the device — the ub table at the open part's centroid foot (RTX_E3_WALL_FOOT=centroid, table only), the solid exchange's wall velocity at the axis feet (RTX_E3_WALL_EXCHANGE=axisfoot: per-face six-entry foot t…

Documentation #987:Commit f6b34f924d pushed by osobh

main

2026-09-20 13:59:40 +00:00

5s

embedded3 S2-7b: the moving-wall forms on the device — the ub table at the open part's centroid foot (RTX_E3_WALL_FOOT=centroid, table only), the solid exchange's wall velocity at the axis feet (RTX_E3_WALL_EXCHANGE=axisfoot: per-face six-entry foot t…

CI #986:Commit f6b34f924d pushed by osobh

main

2026-09-21 14:32:42 +00:00

24h33m9s

embedded3 S2-7b: the moving-wall forms on the device — the ub table at the open part's centroid foot (RTX_E3_WALL_FOOT=centroid, table only), the solid exchange's wall velocity at the axis feet (RTX_E3_WALL_EXCHANGE=axisfoot: per-face six-entry foot t…

Performance Benchmarks #985:Commit f6b34f924d pushed by osobh

main

2026-09-20 14:02:09 +00:00

2m36s

embedded3 S2-7b: the curved instrument's per-wall operator probe (curved_operator_probe) and RTX_E3_CURVED_SCHEME; host prototype RTX_E3_WALL_FOOT=centroid (the shear closure's wall velocity at the foot from the open part's CENTROID, not the face centr…

Documentation #984:Commit 7fffbb1301 pushed by osobh

main

2026-09-20 13:49:49 +00:00

1m3s

embedded3 S2-7b: the curved instrument's per-wall operator probe (curved_operator_probe) and RTX_E3_CURVED_SCHEME; host prototype RTX_E3_WALL_FOOT=centroid (the shear closure's wall velocity at the foot from the open part's CENTROID, not the face centr…

CI #983:Commit 7fffbb1301 pushed by osobh

main

2026-09-21 14:32:42 +00:00

24h43m57s

embedded3 S2-7b: the curved instrument's per-wall operator probe (curved_operator_probe) and RTX_E3_CURVED_SCHEME; host prototype RTX_E3_WALL_FOOT=centroid (the shear closure's wall velocity at the foot from the open part's CENTROID, not the face centr…

Performance Benchmarks #982:Commit 7fffbb1301 pushed by osobh

main

2026-09-20 13:48:50 +00:00

6s