v3.6.0-windows.post25

Fix the bug on AMD GPU: Assertion failed: llvm::isUIntN(BitWidth, val) && "Value is not an N-bit unsigned value" in TritonAMDFoldTrueCmpI, see #184

v3.6.0-windows.post24

There is no change in the Windows support code compared to triton-windows 3.5.1.post24.

Note again that Triton 3.6 only works with PyTorch >= 2.10 .

To install Triton 3.6 and prevent breaking with your installed PyTorch when a new version of Triton is released in future, you need to limit the version to be < 3.7:

pip install -U "triton-windows<3.7"

v3.5.1-windows.post24

Shorten cache temp path to work around Windows' path length limit of 260 chars, see #114 (comment)

v3.5.1-windows.post23

Initial support for AMD GPU with TheRock, see #179

v3.5.1-windows.post22

Fix the warnings Failed to find cuobjdump.exe/nvdisasm.exe, see #157

v3.5.1-windows.post21

There is no change in the Windows support code compared to triton-windows 3.5.0.post21. The only difference between the official triton 3.5.0 and 3.5.1 is a fix for sm103 (GB300) GPUs.

v3.5.0-windows.post21

• Find Windows SDK with some weird installation, see #131
• The patch to support fp8 on RTX 30xx is merged, see #140

Note again that Triton 3.5 only works with PyTorch >= 2.9 .

To install Triton 3.5 and prevent breaking with your installed PyTorch when a new version of Triton is released in future, you need to limit the version to be < 3.6:

pip install -U "triton-windows<3.6"

triton-windows 3.2.0.post21 is also released, which supports fp8 on RTX 20xx and PyTorch 2.6 .

triton-windows 3.3.1.post21 and 3.4.0.post21 are also released, which support PyTorch 2.7 and 2.8 respectively.

v3.4.0-windows.post20

Now Proton profiler's instrumentation backend works. Proton profiler does not fully work yet, so I'm not bundling it in the wheels. If you want to try it, you can build from source.

Note again that Triton 3.4 only works with PyTorch >= 2.8 .

To install Triton 3.4 and prevent breaking with your installed PyTorch when a new version of Triton is released in future, you need to limit the version to be < 3.5:

pip install -U "triton-windows<3.5"

v3.3.1-windows.post19

This is identical with triton-windows 3.3.0.post19, but I bump the version number to match the official one.

The only difference between the official triton 3.3.0 and 3.3.1 is triton-lang#6771 , which is a fix for RTX 50xx GPUs. I've already added this patch since triton-windows 3.3.0.post14.

v3.3.0-windows.post19

• Fix JIT compilation using Clang

Note again that Triton 3.3 only works with PyTorch >= 2.7, and Triton 3.2 only works with PyTorch >= 2.6 .

To install Triton 3.3 and prevent breaking with your installed PyTorch when a new version of Triton is released in future, you need to limit the version to be < 3.4:

pip install -U "triton-windows<3.4"