CoreWeave proves Nvidia's aging AI GPUs from 2020 can generate profit nine years after deployment, signs A100 contracts into 2029

CoreWeave has signed a multi-year contract for Nvidia A100 GPUs extending into 2029, demonstrating that 2020-era Ampere-based chips remain commercially viable nine years after their initial deployment... Read original →

The multi-year A100 leasing deal between CoreWeave and Nvidia reveals a structural imbalance in the AI compute market. Technologically, despite being phased out, A100s remain highly utilized, indicating that current AI workloads prioritize compute availability over performance upgrades. This delays adoption of newer chips like GB200/GB300, which face infrastructure constraints due to power and cooling demands. From a compliance standpoint, the global semiconductor supply chain’s fragmentation increases the scarcity of legacy chips, especially amid U.S.-China tech rivalry. In competitive dynamics, cloud providers like Google may accelerate offerings of legacy-compatible compute to gain market share. Over the next 12–24 months, the AI infrastructure landscape will likely bifurcate into a long-tail model, where older GPUs dominate mainstream use cases, while newer architectures target niche, high-end applications, reshaping the compute competition.