NVIDIA and Shopify use Applied Compute

NVIDIA and Shopify showing up as customers as well as investors means Applied Compute is selling into companies that could have solved this problem internally, but still chose an outside specialist. That points to a product that is not just raw GPU rental. It is a higher touch system for training, serving, and improving custom models on company data, with enough speed and reliability that large technical buyers use it in production.1 2 3

• Applied Compute describes customers including Shopify, Microsoft, DoorDash, Harvey, Cognition, Clay, Mercor, and Decagon using it to build and run AI tailored to their work. That puts Shopify in a broader pattern of buyers using Applied Compute for real workloads, not just experimenting or taking a venture stake.1
• The NVIDIA relationship appears to go beyond cap table signaling. Applied Compute published a case study on post training with NVIDIA Nemotron and said it runs Nemotron across frontier enterprise deployments. That suggests NVIDIA is using Applied Compute both as a channel for enterprise adoption and as an operator for demanding model tuning and inference work.2
• This is different from a plain GPU cloud story. Applied Compute says it helps companies train, deploy, and improve custom AI models on their own data. In the same market, providers like CoreWeave and Lambda are known primarily for selling compute capacity, clusters, and infrastructure tooling. Applied Compute is selling a more packaged outcome.3 4 5

Going forward, customer overlap with strategic investors should make Applied Compute look less like a vendor and more like core AI infrastructure for enterprises. If more large technical companies buy before or alongside investing, the company can compound through deeper deployments, reference accounts, and a tighter loop between model research and production demand.1 2 3