rustytorch

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

Documentation #593:Commit 4f7b350d96 pushed by osobh

main

2026-09-12 02:42:41 +00:00

8s

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

CI #592:Commit 4f7b350d96 pushed by osobh

main

2026-09-13 01:05:47 +00:00

22h23m16s

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

overset added-mass pin (P5-3): a no-slip cylinder on the moving patch oscillating in a closed box of still fluid, the force fitted over the last two periods — C_m = 1.069 (n = 32) / 1.097 (n = 64) against Stokes's 1.064 plus box blockage: the composit…

Documentation #590:Commit 33534ee688 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

overset added-mass pin (P5-3): a no-slip cylinder on the moving patch oscillating in a closed box of still fluid, the force fitted over the last two periods — C_m = 1.069 (n = 32) / 1.097 (n = 64) against Stokes's 1.064 plus box blockage: the composit…

CI #589:Commit 33534ee688 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

overset added-mass pin (P5-3): a no-slip cylinder on the moving patch oscillating in a closed box of still fluid, the force fitted over the last two periods — C_m = 1.069 (n = 32) / 1.097 (n = 64) against Stokes's 1.064 plus box blockage: the composit…

Performance Benchmarks #588:Commit 33534ee688 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

prescribed-motion replay: per-step Schwarz rounds, convergence, background residual, patch divergence and mass defects in the CSV — so the next burst is seen forming

Documentation #587:Commit aeb70eb507 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

prescribed-motion replay: per-step Schwarz rounds, convergence, background residual, patch divergence and mass defects in the CSV — so the next burst is seen forming

CI #586:Commit aeb70eb507 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

prescribed-motion replay: per-step Schwarz rounds, convergence, background residual, patch divergence and mass defects in the CSV — so the next burst is seen forming

Performance Benchmarks #585:Commit aeb70eb507 pushed by osobh

main

2026-09-11 05:08:30 +00:00

0s

prescribed-motion replay: the fluid's power on the flag split face by face (normal vs tangential traction; root / middle / tip thirds) beside the nodal Σ F·ḋ, and replay instants (RTX_FSI2O_SAVE + SAVE_EVERY or SAVE_AT=t1,t2,… — the step that first…

Documentation #584:Commit 1562b7cbb5 pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

prescribed-motion replay: the fluid's power on the flag split face by face (normal vs tangential traction; root / middle / tip thirds) beside the nodal Σ F·ḋ, and replay instants (RTX_FSI2O_SAVE + SAVE_EVERY or SAVE_AT=t1,t2,… — the step that first…

CI #583:Commit 1562b7cbb5 pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

prescribed-motion replay: the fluid's power on the flag split face by face (normal vs tangential traction; root / middle / tip thirds) beside the nodal Σ F·ḋ, and replay instants (RTX_FSI2O_SAVE + SAVE_EVERY or SAVE_AT=t1,t2,… — the step that first…

Performance Benchmarks #582:Commit 1562b7cbb5 pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

P5-3 energy discriminator: prescribed motion on the overset fluid alone — fsi2_harness/replay.rs (a march's saved instants as cubic-Hermite kinematics with the saved velocities as slopes, the prescribed wall velocity the interpolant's exact derivative…

Documentation #581:Commit f61c427b6d pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

P5-3 energy discriminator: prescribed motion on the overset fluid alone — fsi2_harness/replay.rs (a march's saved instants as cubic-Hermite kinematics with the saved velocities as slopes, the prescribed wall velocity the interpolant's exact derivative…

CI #580:Commit f61c427b6d pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

P5-3 energy discriminator: prescribed motion on the overset fluid alone — fsi2_harness/replay.rs (a march's saved instants as cubic-Hermite kinematics with the saved velocities as slopes, the prescribed wall velocity the interpolant's exact derivative…

Performance Benchmarks #579:Commit f61c427b6d pushed by osobh

main

2026-09-10 23:08:30 +00:00

0s

FSI2 overset harness: RTX_FSI2O_DT_SCALE scales the CFL-derived fluid step — P5-3's dt-convergence probe of the first-order Euler fluid steps (the band's placement is excluded: 94.0 mm at a 60 mm patch vs 94.8 at 40 mm; the s = 8 pair moved the amplit…

Documentation #578:Commit 7e13d14e96 pushed by osobh

main

2026-09-10 17:08:30 +00:00

0s

FSI2 overset harness: RTX_FSI2O_DT_SCALE scales the CFL-derived fluid step — P5-3's dt-convergence probe of the first-order Euler fluid steps (the band's placement is excluded: 94.0 mm at a 60 mm patch vs 94.8 at 40 mm; the s = 8 pair moved the amplit…

CI #577:Commit 7e13d14e96 pushed by osobh

main

2026-09-10 17:08:30 +00:00

0s

FSI2 overset harness: RTX_FSI2O_DT_SCALE scales the CFL-derived fluid step — P5-3's dt-convergence probe of the first-order Euler fluid steps (the band's placement is excluded: 94.0 mm at a 60 mm patch vs 94.8 at 40 mm; the s = 8 pair moved the amplit…

Performance Benchmarks #576:Commit 7e13d14e96 pushed by osobh

main

2026-09-10 17:08:30 +00:00

0s

FSI2 overset march: print every coupling setting once (coupler, floor, rtol, stall accept, max subit, predictor, s, patch offset × rows) and add RTX_FSI2O_LOAD_SETTLE (rigid steps after a loaded state, clock put back; default 20) — P5-3 lost a day to …

Documentation #575:Commit 75ab6e77b4 pushed by osobh

main

2026-09-09 23:08:30 +00:00

0s

FSI2 overset march: print every coupling setting once (coupler, floor, rtol, stall accept, max subit, predictor, s, patch offset × rows) and add RTX_FSI2O_LOAD_SETTLE (rigid steps after a loaded state, clock put back; default 20) — P5-3 lost a day to …

CI #574:Commit 75ab6e77b4 pushed by osobh

main

2026-09-09 23:08:30 +00:00

24h45m8s

FSI2 overset march: print every coupling setting once (coupler, floor, rtol, stall accept, max subit, predictor, s, patch offset × rows) and add RTX_FSI2O_LOAD_SETTLE (rigid steps after a loaded state, clock put back; default 20) — P5-3 lost a day to …

Performance Benchmarks #573:Commit 75ab6e77b4 pushed by osobh

main

2026-09-09 23:08:30 +00:00

0s

P5-3 instruments: per-region wall loads on saved instants (top / bottom / tip arc / fillets+cylinder — drag, lift, normal-traction mean/min/max; RTX_FSI2O_AUDIT_FACES, RTX_FSI2O_AUDIT_ONLY) and the patch-thickness knobs (RTX_FSI2O_PATCH_OFFSET / _PATC…

Documentation #572:Commit 3e019d1491 pushed by osobh

main

2026-09-09 17:08:30 +00:00

0s

P5-3 instruments: per-region wall loads on saved instants (top / bottom / tip arc / fillets+cylinder — drag, lift, normal-traction mean/min/max; RTX_FSI2O_AUDIT_FACES, RTX_FSI2O_AUDIT_ONLY) and the patch-thickness knobs (RTX_FSI2O_PATCH_OFFSET / _PATC…

CI #571:Commit 3e019d1491 pushed by osobh

main

2026-09-09 17:08:30 +00:00

23h32m17s

P5-3 instruments: per-region wall loads on saved instants (top / bottom / tip arc / fillets+cylinder — drag, lift, normal-traction mean/min/max; RTX_FSI2O_AUDIT_FACES, RTX_FSI2O_AUDIT_ONLY) and the patch-thickness knobs (RTX_FSI2O_PATCH_OFFSET / _PATC…

Performance Benchmarks #570:Commit 3e019d1491 pushed by osobh

main

2026-09-09 17:08:30 +00:00

0s

P5 audit exact on the moving path: the composite rebuilt around a saved instant carries the march's dt (from the mesh the march started on — the warm base under RTX_FSI2O_WARM_SWEEPS — whatever patch the instant holds; 0.2 % of dt read as 0.3–5 N/m …

Documentation #569:Commit 2c1143076e pushed by osobh

main

2026-09-08 23:08:30 +00:00

0s

P5 audit exact on the moving path: the composite rebuilt around a saved instant carries the march's dt (from the mesh the march started on — the warm base under RTX_FSI2O_WARM_SWEEPS — whatever patch the instant holds; 0.2 % of dt read as 0.3–5 N/m …

CI #568:Commit 2c1143076e pushed by osobh

main

2026-09-08 23:08:30 +00:00

5h33m37s

P5 audit exact on the moving path: the composite rebuilt around a saved instant carries the march's dt (from the mesh the march started on — the warm base under RTX_FSI2O_WARM_SWEEPS — whatever patch the instant holds; 0.2 % of dt read as 0.3–5 N/m …

Performance Benchmarks #567:Commit 2c1143076e pushed by osobh

main

2026-09-08 23:08:30 +00:00

0s

rtx-cfd: the deformed cylinder–flag build's hull source carries the deformed outline — the cylinder + tip disc alone give the chord under a bent flag's convex side (P5-2 died at tip −35 mm with the patch 4.4 h thick where 6 h was asked, an acceptor'…

Documentation #566:Commit 06229dd9a2 pushed by osobh

main

2026-09-08 23:08:30 +00:00

0s

rtx-cfd: the deformed cylinder–flag build's hull source carries the deformed outline — the cylinder + tip disc alone give the chord under a bent flag's convex side (P5-2 died at tip −35 mm with the patch 4.4 h thick where 6 h was asked, an acceptor'…

CI #565:Commit 06229dd9a2 pushed by osobh

main

2026-09-08 23:08:30 +00:00

5h33m37s