A significant development has emerged in the world of GPU acceleration, traditionally dominated by Nvidia. New open-source software now allows AMD GPUs to run native Cuda code, a breakthrough that gives AMD users access to GPU-accelerated AI workloads that were previously exclusive to Nvidia hardware. This development, supported by AMD-funded projects, has the potential to level the playing field in the AI and high-performance computing sectors by expanding software compatibility and empowering a wider range of hardware for AI applications.