The AI industry is running out of compute, with outages, rationing, and rising GPU prices

Editor's take

Demand for AI compute, particularly for training and running large language models, has outstripped current supply, leading to service disruptions and escalating hardware costs. Anthropic has experienced outages, OpenAI has paused public access to its Sora video model, and NVIDIA H100 GPU prices have seen a significant increase, reflecting this imbalance.

This bottleneck directly impacts companies like Anthropic and OpenAI, slowing their ability to deploy advanced models and develop new capabilities. It also signals a potential shift in the AI landscape, where access to cutting-edge compute could become a significant competitive differentiator, potentially favoring well-capitalized players.

Future developments will hinge on NVIDIA's ability to ramp up H100 production and the success of alternative hardware solutions. Monitoring the pricing trends for GPUs and the strategic responses of cloud providers like Microsoft Azure and Amazon Web Services will be crucial to understanding the long-term implications of this compute crunch.