ROCmFix and InferBench are zero-dependency developer utilities built to streamline local AI on consumer AMD GPUs. ROCmFix auto-detects hardware to configure HSA_OVERRIDE_GFX_VERSION overrides across Windows & Linux. InferBench provides controlled, isolated benchmarking for Ollama and LM Studio APIs to measure Vulkan vs HIP inference speeds.
Hey Product Hunt! I built ROCmFix and InferBench to solve the configuration friction and backend performance debate when running local AI models on Radeon hardware. Both scripts are zero-dependency, open-source Python tools. Excited to hear your feedback and see your benchmark comparisons!