Quantitative trading firm Hudson River Trading (HRT) has entered a multi-year agreement with CoreWeave (NASDAQ: CRWV) to access specialized artificial intelligence computing resources. The arrangement, described as multi-billion-dollar in scale, will support the development of next-generation trading research and machine learning models.
It builds on an earlier relationship that began when HRT first became a CoreWeave customer in March 2026, representing a substantial expansion of their collaboration.
Under the deal, HRT will utilize CoreWeave’s purpose-built AI cloud platform.
This includes clusters of NVIDIA Vera Rubin NVL72 systems and NVIDIA HGX B200 GPU systems, designed for the high density and throughput required in large-scale model training.
The infrastructure also features high-throughput Direct Connect networking based on NVIDIA Spectrum-X Ethernet technology.
This setup aims to deliver dedicated, high-bandwidth connectivity between HRT’s on-premises systems and CoreWeave’s cloud environment, reducing latency and enabling more predictable data transfers.
HRT, recognized as one of the world’s leading quantitative trading firms and among the largest nonbank trading operations in the United States, has long applied machine learning, neural networks, and high-performance computing to its research and trading activities.
As AI plays a larger role in quantitative strategies, the firm requires scalable, high-performance infrastructure to train increasingly complex models, handle expanding data volumes, and accelerate iteration cycles that keep pace with fast-moving markets.
Kevin Lee, head of Research & Development at HRT, highlighted the strategic importance of the platform choice.
He noted that the foundation supporting AI and machine learning research is as critical as the models themselves.
Lee explained that CoreWeave was selected for its understanding of the demands of running AI in rigorous production settings and its capacity to grow in step with HRT’s goals.
Access to the newest AI technologies, he added, allows research teams to advance more effectively and rapidly.Jon Jones, chief revenue officer at CoreWeave, emphasized the significance of the financial services sector as a challenging test bed for AI applications.
He described HRT as a company that exemplifies bringing AI into high-stakes, performance-sensitive production environments.
Jones stated that the partnership underscores the confidence sophisticated organizations place in CoreWeave’s ability to handle demanding AI workloads.
The agreement positions HRT among the early CoreWeave customers gaining broad access to NVIDIA’s latest Vera Rubin AI chips.
CoreWeave has been expanding its footprint in financial services, adding other quantitative firms as clients while diversifying beyond its traditional technology and AI laboratory customer base.
Competitors such as Jane Street have also deepened ties with CoreWeave through substantial capacity commitments and investments, reflecting broader industry demand for specialized AI compute capacity.
This collaboration enables HRT researchers to experiment with greater flexibility through direct technical support and a range of tools and frameworks.
It supports the firm’s ongoing efforts to innovate in multi-asset class trading, liquidity provision, and risk management across global markets.
For CoreWeave, the deal reinforces its strategy of serving organizations that operate AI at production scale and demonstrates its infrastructure performance credentials, including strong benchmark results in industry evaluations.
As quantitative trading firms continue to integrate advanced AI into their core operations, partnerships like this one between HRT and CoreWeave illustrate the growing reliance on specialized cloud providers capable of delivering hardware and reliable, low-latency connectivity. The multi-year commitment signals sustained investment in computational power as a competitive differentiator in high-frequency and AI-driven trading environments.