Infrastructure
CUDA
The software platform from NVIDIA that lets general code run on its GPUs, and a big reason NVIDIA dominates AI hardware.
CUDA is the programming layer that lets developers harness the thousands of parallel cores in an NVIDIA GPU for general computation, including the math behind neural networks. Because nearly all major AI frameworks are built on it, CUDA has become a deep moat: it ties the ecosystem to NVIDIA hardware and is a large part of why the company leads the AI chip market. It is the bridge between AI software and the GPU underneath.
Reference: Wikipedia: CUDA
Related terms
Graphics Processing Unit
The specialized chip that does the heavy parallel math behind training and running AI models.
Compute
The raw processing power used to train or run a model, a key driver of cost and capability.
Tensor Processing Unit
A chip designed by Google specifically for machine learning, as an alternative to general-purpose GPUs.