ROCm

AMD's open-source counterpart to CUDACUDA, prop. n.: Nvidia's software system for programming its graphics chips to do general calculations. Most AI software is built on it, which is a big reason Nvidia chips dominate AI., the software that lets programs use its Instinct and Radeon GPUs for general computation. It bundles drivers and runtime, compilers, math and AI libraries (such as rocBLAS, MIOpen and the RCCL communication library), profilers and HIP, a C++ programming interface closely modeled on CUDA, with HIPIFY tools that translate CUDA source code into HIP. PyTorchPyTorch, prop. n.: The most widely used free software toolkit for building and training neural networks, especially in research; programmers write ordinary Python and it handles the calculus and the graphics chips., JAXJAX, prop. n.: A free Google library that lets researchers write mathematics in familiar Python and automatically get derivatives and very fast code for graphics chips and Google's AI chips., TritonTriton, prop. n.: A free language from OpenAI for writing the fast inner routines of AI programs in Python-style code; its compiler handles hardware details that Nvidia's own tools leave to experts. and inference engines such as vLLMvLLM, prop. n.: A free, open-source program for running large language models on servers efficiently, so one machine can answer many more users at once. Many companies use it to host AI models. support ROCm, so much CUDA-targeted AI software can run on AMD accelerators such as the MI300X (2023) and MI350 series (2025). Its historic weaknesses relative to CUDA — narrower hardware support, rougher tooling and fewer tuned kernels — have, by many accounts, narrowed since 2024 as large buyers deployed AMD GPUs for inference and training, though CUDA's ecosystem remains broader.