perf(prover): per-table scheduler with VRAM admission for multi_prove - #877

Draft

ColoCarletti wants to merge 38 commits into

mainfrom

gpu-opt-table-scheduler

Commits on Jul 23, 2026

add profiling
ColoCarletti
committed

Commits on Jul 24, 2026

fix(profiling): field fixes from the first sessions on the 5090 box
ColoCarletti
committed
docs(profiling): complete the toolkit README as a reference
ColoCarletti
committed
perf(gpu): async pinned D2H + pre-created event pool + precomputed-tree cache
ColoCarletti
committed
perf(prover): pipeline + concurrent epoch proving in continuations
ColoCarletti
committed
perf(gpu): dim-split constraint interpreter with liveness-reused value slots
ColoCarletti
committed
perf(gpu): commit preprocessed tables through the fused GPU pipeline
ColoCarletti
committed
perf(prover): overlap the global prove with the epoch proves' tail
ColoCarletti
committed
perf(prover): share per-ELF DECODE artifacts across continuation epochs
ColoCarletti
committed
perf(prover): pipeline epoch trace builds onto a builder pool
ColoCarletti
committed
perf(prover): cache pre-captured AIR prototypes per table type
ColoCarletti
committed
profiling: re-home the toolkit spans onto the pipelined continuation flow
ColoCarletti
committed

Commits on Jul 27, 2026

Commits on Jul 28, 2026

Merge branch 'main' into gpu-opt-5090
diegokingston
authored

Commits on Jul 29, 2026