dtourolleandClaude Opus 5
e596eb0657
The scan is O(n²) dot products and nothing else, so its speed is the face subsystem's speed — and it was running at 0.7 flops per cycle. Two separate faults, both measured over the reference 18,143-face library on twenty cores. It walked the whole embedding array once per row, ~336 GB of traffic, where a column tile that fits in L2 is read once per tile of rows: 4.64s → 2.81s. And the workspace builds for baseline x86-64 — SSE2, no FMA — into which the portable loop was not being vectorised at all: 2.81s → 0.86s, 195 GFLOP/s. So the dot product is now chosen per machine. AVX2 + FMA where is_x86_feature_detected! finds it; NEON unconditionally on aarch64, since Advanced SIMD is in that baseline and every Android device the app builds for has it — with the explicit vfmaq, because LLVM will not fuse a multiply and an add without being told to. The portable loop stays as the definition the others are tested against, and the_fastest_kernel_agrees_with_the_portable_one is the only check the NEON path gets on a machine that is not aarch64. Faces::embeddings is one flat buffer rather than a Vec per face: the pointer chase defeated both the prefetcher and the tiling, and it is also the layout a GPU pass would want. Behaviour is unchanged and that is checked rather than asserted — the same 1,531,969 pairs from all three kernels, and on the real library the same 2,518 groups holding the same 16,246 faces with the same confidence distribution. A full regroup there goes from 10.0s to 5.9s; the rest is the agglomeration, which is a sequential heap walk and is where the next look should go. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
DarkRoom
A cross-platform, non-destructive RAW photo editor for Linux and Android.
Status: early. v0.1 is a remote library viewer — see docs/milestone-v0.1.md.
Documentation
| Document | Contents |
|---|---|
| requirements.md | What the software must do — 122 numbered requirements |
| architecture.md | How it is built — crates, GPU pipeline, data model, sync |
| milestone-v0.1.md | The first buildable milestone |
| faces.md | Face detection and identity — the models, the licence problem, and what S14 measures |
Building
Desktop:
cargo run -p darkroom-desktop
Android (containerised toolchain, see docker/android):
./docker/android/build.sh cargo ndk -t arm64-v8a build --release
Current state
Working: workspace, GPU context and compute pass, adaptive Slint shell, Android cross-compilation of the core crates.
Not yet working: the zero-copy display path. The build currently uploads frames through the CPU, which is exactly what ARCH §6.1 forbids — measured at 96% of frame time at 4K. Replacing it is spike S1, the project's highest priority.
cargo run -p dr-gpu --example bench --features readback
reproduces that measurement.
Licence
GPL-3.0-or-later.