v0.11.0 — Strix Halo, MI210, MI300X

Tagged v0.11.0. h3 --info prints h3-hip 0.11.0.

Timed SKUs: Strix Halo (gfx1151), MI210 (gfx90a), MI300X (gfx942).
MI250 / MI250X share gfx90a; they were not timed.

Quality path (default)

Preset	Strix Halo (gfx1151)	MI210 (gfx90a)	MI300X (gfx942)
fox-s2 E2E	~85–90 s (I/O)	~10.8 s	~16 s
fox-fast E2E	~2 min (I/O)	~18 s	~12 s
15 s cinematic E2E	40 min 46 s	12 min 11 s	3 min 46 s

Strix Halo vs v0.9.0: fox-s2 denoise 6.3–6.5 s → 3.36 s; 15 s E2E 45.0 min → 40 min 46 s.

Speed path (opt-in)

--token-reduction + H3_INT8_VAE=1 (CDNA also H3_INT8_MLP=1 in all-opts). Visible quality trade; not the gallery clip.

15 s cinematic	Strix Halo (gfx1151)	MI210 (gfx90a)	MI300X (gfx942)
Speed-path E2E	27 min 3 s	8 min 21 s	~2.4 min

Build

git clone https://github.com/alexhegit/h3-hip.c.git
cd h3-hip.c
git checkout v0.11.0
make HIP_ARCH=gfx1151 -j$(nproc) h3   # or gfx90a / gfx942

Scoreboard: PERFORMANCE.md · project page · wiki

v0.10.1 — docs patch on the dual-ISA line

Docs patch on v0.10.0. Same kernels and scoreboard. h3 --info prints h3-hip 0.10.1.

fox-s2 and fox-fast are complete muxed MP4s (512² · 22 frames, ~0.9 s), not stubs. The README fox showcase clip is a third preset (--layers 50 --reuse 1). The 15 s cinematic reuses fox-fast knobs at 864×480.

git clone https://github.com/alexhegit/h3-hip.c.git
cd h3-hip.c
git checkout v0.10.1
make HIP_ARCH=gfx1151 -j$(nproc) h3   # or gfx90a

Preset	gfx1151	gfx90a
fox-s2 E2E	83–87 s	~10.5 s
fox-fast E2E	~95 s	~18 s
15 s cinematic E2E	45.0 min	12 min 33 s

gfx1151 fox-s2 md5 unchanged from v0.9.0: 1731f95c4aa582597cf83d57f46b8f9e.

Docs: Getting started · Performance

v0.10.0 — gfx1151 and gfx90a

One HIP tree, two ISAs. Pass HIP_ARCH yourself; the Makefile does not probe the GPU.

ISA	GPU	Default DiT	SDPA
gfx1151	Strix Halo	INT8 + BF16 activations	rocWMMA
gfx90a	MI210	BF16 GEMM	MFMA flash

Build:

git clone https://github.com/alexhegit/h3-hip.c.git
cd h3-hip.c
git checkout v0.10.0
make HIP_ARCH=gfx1151 -j$(nproc) h3   # or gfx90a

Preset	gfx1151	gfx90a
fox-s2 E2E	83–87 s	~10.5 s
fox-fast E2E	~95 s	~18 s
15 s cinematic E2E	45.0 min	12 min 33 s

gfx1151 fox-s2 md5 unchanged from v0.9.0: 1731f95c4aa582597cf83d57f46b8f9e.

Docs: Getting started · Performance