rustytorch

PERF-2 P1.1: the multigrid-PCG's prepared operator (hierarchy, f64 fine level, active cells, components) cached on the embedded solver and reused while the operator is bit-identical (an exact key over the coefficient bit patterns, the mask and the hier…

Performance Benchmarks #658:Commit 79c18def32 pushed by osobh

main

2026-09-16 04:34:04 +00:00

3m7s

PERF-2 P1: exact chunked polyline index for the O-grid's sliding outer ring — PolylineIndex::nearest returns nearest_on_polyline's point bit for bit (same segments, same order, same strict minimum; chunks skipped only when their box is farther than th…

Performance Benchmarks #655:Commit 01830e5c8d pushed by osobh

main

2026-09-16 04:12:03 +00:00

3m48s

rtx-cfd: the regeneration profile block sits above the too_many_arguments attribute, not between it and its function

Performance Benchmarks #652:Commit 11f71b832c pushed by osobh

main

2026-09-16 03:57:55 +00:00

4m12s

PERF-2 P0-b: CG iterations per Poisson solve in the profile (ms per V-cycle) and sub-timers inside the patch regeneration (outline, hull + offset, ring projection, Winslow sweeps, respace, warm bookkeeping, mesh finalisation)

Performance Benchmarks #649:Commit 1547d14ff1 pushed by osobh

main

2026-09-16 03:57:33 +00:00

4m4s

rtx-fsi harness: 'fixed' coupler (constant relaxation OMEGA0) for the P6-c Robin–Neumann test — at ω = 1 the candidate is the FEA's own output, so the Robin wall's previous-pass datum is the structure's traction at the candidate; design 2's FSI3 div…

Performance Benchmarks #643:Commit 2d2fed7181 pushed by osobh

main

2026-09-16 00:27:59 +00:00

5m19s

P6-b design 2: the Robin wall on the patch's Inner side — RobinWall { alpha, datum } with the wall velocity u_s + (t_f − datum)/α (explicit, damped by the wall's own viscous gain μ/(α d)) and the compliant-wall pressure term |S| p'/α implicit in …

Performance Benchmarks #634:Commit c3dbd5040a pushed by osobh

main

2026-09-15 15:49:36 +00:00

2m55s

P6-b: the compensating load uses the candidate's own Newmark acceleration a(d) = (d − u_pred)/(β Δt²) (consistent with the relaxed candidate the fluid saw; cancels exactly at the fixed point) — the lagged structure-output acceleration of the first …

Performance Benchmarks #631:Commit c144a5f733 pushed by osobh

main

2026-09-15 13:40:14 +00:00

2m37s

fsi2 harness: RTX_FSI2O_PATCH_CONVECTION knob (tvd default / upwind / none) on the overset patch, printed in the march's coupling header and the replay's PRESCRIBED header — P5-3's near-wake dispersion test on the ny 62 fixed-motion replay; harness fi…

Performance Benchmarks #604:Commit ff7cc2d739 pushed by osobh

main

2026-09-13 02:21:04 +00:00

2m58s

rtx-backend-cuda: native index_select / index_add — HetGAT training 2.5x, inference 4.3x

GPU Tests #597:Commit 796b8487ad pushed by osobh

main

2026-09-12 04:09:08 +00:00

2s

rtx-backend-cuda: native index_select / index_add — HetGAT training 2.5x, inference 4.3x

Performance Benchmarks #594:Commit 796b8487ad pushed by osobh

main

2026-09-12 04:12:08 +00:00

3m5s

rtx-backend-cuda: transpose of tall tensors failed to launch, and the generic permute was an identity copy

Performance Benchmarks #591:Commit 4f7b350d96 pushed by osobh

main

2026-09-12 02:49:53 +00:00

7m22s

feat(mamba): GPU-accelerated backward pass (backward_cuda)

GPU Tests #278:Commit 5155c081ca pushed by osobh

main

2026-08-10 14:11:31 +00:00

3s

feat(mamba): GPU-accelerated backward pass (backward_cuda)

Performance Benchmarks #275:Commit 5155c081ca pushed by osobh

main

2026-08-10 14:17:23 +00:00

5m56s

fix(streaming): sane DynamicBatchingConfig default; worker lifecycle regression tests

Performance Benchmarks #272:Commit ad6405663f pushed by osobh

main

2026-07-11 00:16:33 +00:00

26s

fix(streaming): real inference backend wiring and lifecycle fixes; full suite green

Performance Benchmarks #266:Commit a0bf29461b pushed by osobh

main

2026-07-10 22:49:21 +00:00

45s

docs: JEPA roadmap — GPU resume re-upload done; remaining items need multi-GPU

Performance Benchmarks #260:Commit fce6cef262 pushed by osobh

main

2026-07-10 11:00:16 +00:00

26s

feat(jepa): GPU weight re-upload on checkpoint resume

Performance Benchmarks #257:Commit a755627269 pushed by osobh

main

2026-07-10 11:00:49 +00:00

1m10s

feat(jepa): eval in the training loop, NCCL GPU AllReduce, GPU checkpointing

Performance Benchmarks #248:Commit ac3f2af06b pushed by osobh

main

2026-07-10 10:52:08 +00:00

29s

feat(jepa): extended GPU training, data pipeline, integration, and cargo config

Performance Benchmarks #242:Commit 19b6581f9c pushed by osobh

main

2026-07-10 07:04:52 +00:00

26s

docs(consolidation): audit outcome — flagged MoE/flash-attn duplicates are not duplicates

Performance Benchmarks #236:Commit 4ba1b78215 pushed by osobh

main

2026-07-10 05:31:59 +00:00

1m22s

feat(jepa): full GPU-resident ViT block on CUDA

Performance Benchmarks #233:Commit fdc1432072 pushed by osobh

main

2026-07-10 05:14:37 +00:00

4m6s

feat(demos,inference): wire simulation demos to real compute; fix embedding lookup and weight-name aliases

GPU Tests #232:Commit e080748d88 pushed by osobh

main

2026-07-10 05:06:39 +00:00

6s

feat(inference): concrete EAGLE draft model + real tokenizer at the serving boundary

GPU Tests #228:Commit 733b02cd8b pushed by osobh

main

2026-07-10 04:54:21 +00:00

8s

feat(inference): concrete EAGLE draft model + real tokenizer at the serving boundary

Performance Benchmarks #225:Commit 733b02cd8b pushed by osobh

main

2026-07-10 04:54:36 +00:00

28s

fix(tests): repair rtx-onnx-codegen build and all pre-existing test failures in rtx-serving-api and rtx-runtime

Performance Benchmarks #222:Commit 0cbfc1a739 pushed by osobh

main

2026-07-10 02:49:36 +00:00

31s

chore(sweep): delete 43 orphaned source files; document SYCL/demo/duplication status

GPU Tests #221:Commit 5f32165184 pushed by osobh

main

2026-07-10 02:35:29 +00:00

4s

fix(deps): bump candle-core/nn/transformers 0.8->0.11 for CUDA 13.1 build

GPU Tests #217:Commit 522400a72b pushed by osobh

main

2026-07-10 01:10:53 +00:00

7s

feat(jepa): extended GPU training, data pipeline, integration, and cargo config

Performance Benchmarks #211:Commit e1b4061c23 pushed by osobh

main

2026-06-29 21:36:22 +00:00

43s

fix(rtx-science): drop unused ndarray-linalg dep

Performance Benchmarks #208:Commit b861b3bb2e pushed by osobh

main

2026-06-27 18:01:48 +00:00

2m1s

fix(demos): drop openblas/metal from default features (CPU-only default, backends opt-in)

Performance Benchmarks #205:Commit 70da8215a2 pushed by osobh

main

2026-06-27 16:19:54 +00:00

4m28s