Whoever has the best computers will determine the future of artificial intelligence.
Nvidia has used its experience developing graphics processing units (GPUs) — which were originally developed for applications like gaming but turned out to be rather good at AI tasks — to become a dominant force in AI infrastructure and applications.
But a host of established companies and startups are investing heavily to capture a share of the growing AI computing market. This includes building new types of chips specifically for training and running AI models, evolving traditional processors like CPUs and GPUs to better support AI workloads, and exploring novel information processing technologies such as quantum and neuromorphic computing.
In the market map below, we identify 234 companies building AI computing hardware solutions across 20 categories.
Please click to enlarge.
