Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Only auto-enable the ROCm comfy-kitchen Triton backend on matrix-core GPUs (#14869)

liminfei-amd1377a2f72925
diff --git a/comfy/cli_args.py b/comfy/cli_args.py
index 0d7df5e13..e2e0d97ec 100644
--- a/comfy/cli_args.py
+++ b/comfy/cli_args.py
@@ -92,6 +92,7 @@ parser.add_argument("--directml", type=int, nargs="?", metavar="DIRECTML_DEVICE"
parser.add_argument("--oneapi-device-selector", type=str, default=None, metavar="SELECTOR_STRING", help="Sets the oneAPI device(s) this instance will use.")
parser.add_argument("--supports-fp8-compute", action="store_true", help="ComfyUI will act like if the device supports fp8 compute.")
parser.add_argument("--enable-triton-backend", action="store_true", help="ComfyUI will enable the use of Triton backend in comfy-kitchen. Is disabled at launch by default.")
+parser.add_argument("--disable-triton-backend", action="store_true", help="Force-disable the comfy-kitchen Triton backend, overriding the automatic ROCm/AMD default and --enable-triton-backend.")
class LatentPreviewMethod(enum.Enum):
NoPreviews = "none"
diff --git a/comfy/quant_ops.py b/comfy/quant_ops.py
index 91b3e4fe9..b1aabdc93 100644
--- a/comfy/quant_ops.py
+++ b/comfy/quant_ops.py
@@ -3,6 +3,22 @@ import logging
from comfy.cli_args import args
+
+def _rocm_kitchen_arch_supported():
+ """comfy-kitchen's INT8 Triton kernels compile tl.dot to matrix-core instructions.
+ RDNA3/3.5/4 (gfx11xx/gfx12xx) have WMMA and CDNA (gfx9xx) has MFMA; RDNA1/RDNA2
+ (gfx10xx) have neither, so the INT8 path hangs the GPU there. Gates the automatic
+ ROCm default so those cards stay on the eager fallback (an explicit
+ --enable-triton-backend still forces it on any arch)."""
+ try:
+ arch = torch.cuda.get_device_properties(torch.cuda.current_device()).gcnArchName.split(":")[0]
+ except Exception:
+ return False
+ if arch.startswith(("gfx11", "gfx12")):
+ return True
+ return arch in ("gfx908", "gfx90a", "gfx940", "gfx941", "gfx942", "gfx950")
+
+
try:
import comfy_kitchen as ck
from comfy_kitchen.tensor import (
@@ -26,9 +42,13 @@ try:
logging.warning("WARNING: You need pytorch with cu130 or higher to use optimized CUDA operations.")
# On ROCm/AMD the CUDA backend is unavailable, so Triton is the only accelerated
- # comfy-kitchen backend. Enable it by default there, but only on Triton >= 3.7:
+ # comfy-kitchen backend. Enable it by default there, but only on Triton >= 3.7 AND a
+ # matrix-core GPU (RDNA3+ WMMA gfx11xx/gfx12xx, CDNA MFMA gfx9xx). RDNA1/RDNA2
+ # (gfx10xx) have no WMMA -> the INT8 tl.dot path hangs the GPU, so they stay eager.
# older Triton lacks libdevice.rint on the HIP backend and hard-crashes the INT8 path.
- if args.enable_triton_backend or torch.version.hip is not None:
+ if args.disable_triton_backend:
+ ck.registry.disable("triton")
+ elif args.enable_triton_backend or (torch.version.hip is not None and _rocm_kitchen_arch_supported()):
try:
import triton
triton_version = tuple(int(v) for v in triton.__version__.split(".")[:2])