v0.3.1

MiniDXNN v0.3.1

Minor update: bumps the Agility SDK / DXC baseline and adopts InterlockedAccumulate for vector accumulation.

Changes from v0.3.0

• Updated Agility SDK to 1.721-preview, DXC to v1.10.2605.4
• Use InterlockedAccumulate for vector accumulation
• Fixed feature-support check in unit tests
• Documentation updates

Requirements

• Windows 11 with Developer Mode
• GPU supports Shader Model 6.10 and LinAlg Matrix in D3D12 (AMD Radeon™ RX 9000 Series GPUs or equivalent NVIDIA)
• CMake 3.21+, Visual Studio 2022 (C++20)
• Agility SDK 1.721-preview, DXC v1.10.2605.4
• Python 3.8+ with PyTorch (optional, for reference training)

License

MIT — Copyright (c) 2026 Advanced Micro Devices, Inc.