Training and running AI models takes enormous computing power. This category covers the hardware and infrastructure behind it: GPUs and custom AI chips, data center buildouts, cloud capacity, energy consumption and new ideas for scaling compute. Understand who controls the resources that shape the pace of AI progress.