CoreWeave has signed a multi-year agreement to provide Hudson River Trading with AI cloud infrastructure for its quantitative research and model-development work.
The trading firm plans to use NVIDIA Vera Rubin NVL72 and HGX B200 GPU systems through CoreWeave’s platform, alongside Spectrum-X Ethernet networking. The companies said the deployment is intended to support larger data volumes, more complex machine-learning models and faster research cycles.
The arrangement also includes a dedicated Direct Connect design linking HRT’s on-premises environment with CoreWeave’s cloud infrastructure. That connection is intended to give HRT high-bandwidth, lower-latency access to the computing environment used for AI research.
HRT uses machine learning, neural networks and high-performance computing in quantitative trading. For that type of workload, infrastructure decisions extend beyond GPU availability. Network design, data movement and predictable performance can affect how quickly researchers can train, test and refine models.
“As we scale our AI and machine learning research, the AI platform we build on matters as much as the models we build,” Kevin Lee, HRT’s head of research and development, said in the announcement.
The agreement illustrates a growing use of specialized AI cloud platforms in financial-services research. The practical significance is the move toward integrated compute and networking environments for organizations whose AI workloads are both data-intensive and sensitive to performance consistency.