jgrusewski c2edb273ca perf: batch_size 16384 → 8192 — 2x faster GEMMs, more frequent updates ...

At 16384, each GEMM saturates 97% of 132 SMs — no room for multi-stream
parallelism. At 8192, GEMMs use ~50% SMs, allowing branch streams and
aux parallelism to fill idle SMs. 2x more steps/epoch but each ~2x faster.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-19 09:01:34 +02:00

..

a100.toml

fix: restore explicit buffer_size in GPU profiles (auto-sizing removed)

2026-04-04 01:24:19 +02:00

default.toml

fix: restore explicit buffer_size in GPU profiles (auto-sizing removed)

2026-04-04 01:24:19 +02:00

h100.toml

perf: batch_size 16384 → 8192 — 2x faster GEMMs, more frequent updates

2026-04-19 09:01:34 +02:00

rtx3050.toml

fix: restore explicit buffer_size in GPU profiles (auto-sizing removed)

2026-04-04 01:24:19 +02:00