According to Decrypt, neoclouds such as CoreWeave and Nebius offer cloud computing services that rent Nvidia GPUs to artificial intelligence research labs. Several of these providers began operations by mining cryptocurrency before transitioning or expanding into AI-focused cloud offerings. The article explains how the neocloud business model functions in supporting AI development.
These cloud platforms supply GPU infrastructure on a rental basis, allowing AI laboratories to access high-performance graphics processing units without maintaining large in‑house hardware fleets. The service model typically involves providers managing data centers, networking, and maintenance while customers pay for compute time or dedicated instances as needed. Some neocloud operators previously engaged in crypto mining, which informed aspects of their early infrastructure choices and operational focus.
The shift toward AI workloads reflects broader demand for scalable, on‑demand GPU resources in machine learning research, model training, and inference tasks. By offering rented Nvidia GPUs, neoclouds aim to reduce barriers for labs seeking powerful computational capacity without significant upfront capital investment.
