PTX generation for NVIDIA CUDA GPUs with automatic compute capability detection SPIR-V generation for cross-vendor GPUs (Intel, AMD, NVIDIA, ARM) via OpenCL/Vulkan This library is optimized for ...

TorchInductor is a new compiler backend that compiles FX Graphs generated by TorchDynamo into optimized C++/Triton kernels. This tutorial will guide you through the process of using TorchInductor on a ...

Abstract: The past few years, traditional compiler optimization methods have been found to be further enhanced by machine learning (ML), deep learning (DL) and reinforcement learning (RL). These ...

Abstract: This paper presents ACPO: An AI-Enabled Compiler Framework; a novel framework that provides LLVM with simple and comprehensive tools to enable employing ML models for different optimization ...