What it is
Compute, in the realm of artificial intelligence, specifically denotes the computational resources necessary for AI workloads. This primarily involves processing units like Graphics Processing Units (GPUs), but also includes Central Processing Units (CPUs) and Application-Specific Integrated Circuits (ASICs). These resources are essential for both model training, where vast amounts of data are processed to teach the AI, and inference, where the trained model applies its knowledge to new data to generate outputs.
The demand for compute is a major driver of investment in the AI sector, particularly for hardware manufacturers and cloud computing providers. Companies like Nvidia, which produce high-performance GPUs, see significant revenue growth from this demand. The cost and availability of compute influence the pace of AI development, the competitive landscape for AI startups, and the profitability of cloud services. Monitoring compute capacity and pricing is key to understanding the AI supply chain.
Why it matters
Compute is the fundamental resource for AI, driving demand for specialized hardware and cloud services, impacting tech sector investments.
Reviewed under editorial standardsUpdated September 26, 2026Not investment advice