Compute
Shorthand for the raw computing power — chips, and the time spent running them — needed to train and run AI. Compute is one of the field's scarcest and most strategic resources, and the reason specialised chips and data centres matter so much.
When people in AI talk about “compute”, they mean the sheer processing power a task demands: the specialised chips it runs on, and the hours or weeks of running them. Training a large model is one of the most compute-hungry things humans currently do, which is why the biggest models cost so much and take so long to build. Every answer a deployed model gives — its inference — uses compute too, just on a smaller scale.
The workhorse chip has been the GPU, a processor originally designed for video-game graphics that turned out to be well suited to the maths behind neural networks. Because these chips are expensive and in high demand, access to them has become a genuine strategic concern, and a wave of companies — several of them British, such as Graphcore, Lumai, and Fractile — are designing new kinds of AI accelerator to do the job faster or more efficiently.
Compute has also become a matter of national policy. The UK’s push for “sovereign” AI capacity, and firms like Nscale building data centres to supply it, are really about one thing: making sure the country has enough compute to train and run AI without depending entirely on others.