TRELLIS.2 INT8 ConvRot now runs natively in ComfyUI on AMD ROCm
DrBearJ3w · reddit · 2026-07-28
A developer released a patch kit to run TRELLIS.2’s INT8 ConvRot checkpoint natively in ComfyUI on AMD ROCm, including a ready-made 1024 workflow.
Key points:
- Uses fused W8A8 Triton kernels instead of dequantizing GGUF weights on every forward pass.
- Supports native INT8 .safetensors loading, not GGUF.
- Adds 512, 1024, and 1024-cascade routing, plus a dedicated ComfyUI load node.
- Aims to let one ComfyUI process handle TRELLIS, Krea2, Blender integration, and GLSLShader workflows.
The repo also includes checkpoint verification and reproducible local-build scripts.
More from Infra
- MLA-based KV cache costs 12 GB per million tokens, with KDA state at 230 MB BF16 — zephyr_z9 · 2026-07-28
- A 70B GGUF model stalls on an AMD R9700 as VRAM hits 30 GB but RAM stays flat — Developer-Y · 2026-07-28
- Reddit debates the cheapest practical way to run K3 locally — ylchao · 2026-07-28
- K3 hit a usage-limit exploit on Higgsfield within a day of launch — Mediocre-Witness-778 · 2026-07-28
- AI data-center buildout is reshaping grid incentives and reserve power — kleffew94 · 2026-07-28
- The Verge says Moonshot’s open Kimi K3 could undercut closed U.S. AI models — The Verge AI · 2026-07-28