rob thijssen 1fea51cfc3 engine-cuda: loop kernel takes the diagonal from the launch parameters ...

MiningUniforms carries the twelve diagonal constants (416 bytes now) and
the loop kernel reads them there (LAIR_LOOP_PARAM_DIAG=1) instead of from
the __constant__ table. Static: still 13 loads per internal round (ptxas
reloads either way) but spill 88 -> 32 bytes. Measured, three interleaved
rounds, loop kernel only: beast 2135 -> 2162 MH/s (+1.3%), benjy 725 ->
730 (+0.7%); parity 300/300 and 200/200. The unrolled kernel is byte-
identical (21,120 instructions, 110 registers).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ue5ZZm1Hiv5zPnucykKKuF

2026-09-14 19:51:22 +03:00

..

bench-harness

bench: record the memory clock alongside the SM clock (schema 2)

2026-09-03 18:47:17 +03:00

engine-cpu

Fix windows again (#72)

2026-06-30 16:30:11 +08:00

engine-cuda

engine-cuda: loop kernel takes the diagonal from the launch parameters

2026-09-14 19:51:22 +03:00

engine-gpu

metrics: stale work is the batch in flight at cancellation, not the whole search

2026-09-03 14:23:37 +03:00

metrics

metrics: stale work is the batch in flight at cancellation, not the whole search

2026-09-03 14:23:37 +03:00

miner-cli

engine-cuda: native CUDA mining engine behind MinerEngine

2026-09-03 14:52:10 +03:00

miner-service

engine-cuda: native CUDA mining engine behind MinerEngine

2026-09-03 14:52:10 +03:00

miner-telemetry

Address v12 issues (#65)

2026-06-12 10:48:01 +08:00

pool-service

Send miner auth token in Ready handshake. (#78)

2026-08-13 09:50:03 +08:00

pow-core

Optimize GPU Poseidon2 mining 4.1x on Apple Silicon (u64 shader, midstate, lazy squeeze)

2026-08-11 23:40:12 +08:00

quic-transport

Send miner auth token in Ready handshake. (#78)

2026-08-13 09:50:03 +08:00

solver-wasm

Add captcha share pool service and browser WASM solver (#74)

2026-07-07 17:52:38 +08:00