Wall Street’s AI computing race is turning into a multibillion-dollar payday for CoreWeave.
Hudson River Trading, one of the largest nonbank trading firms in the US, has expanded its relationship with CoreWeave through a multiyear agreement worth billions of dollars, Bloomberg reported. The deal expands an existing relationship between the two companies, with HRT set to use CoreWeave’s AI cloud infrastructure, including Nvidia’s new Vera Rubin NVL72 systems and HGX B200 GPU platforms.
Financial terms were not disclosed. CoreWeave Chief Revenue Officer Jon Jones described the agreement as “a material expansion” in an interview with Bloomberg.
HRT will use the infrastructure to train and test increasingly complex AI models, while moving large volumes of data between its own systems and CoreWeave through dedicated Spectrum-X Ethernet networking, according to CoreWeave.
“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 & Development, said in CoreWeave’s announcement.
The companies say the infrastructure is designed to give HRT researchers more computing capacity, faster data movement and access to newer AI hardware.
The race for compute
The agreement puts HRT among the first CoreWeave customers expected to receive wide-scale access to Nvidia’s Vera Rubin systems, Bloomberg reported. That matters in quantitative trading, where researchers use large datasets and increasingly sophisticated models to test strategies and identify patterns.
“Having access to more compute resources just allows our researchers to try more things and hopefully find more interesting results,” Lee told Bloomberg.
CoreWeave is also pushing deeper into financial services as it tries to reduce its dependence on large technology customers. Its largest customer accounted for 36% of second-quarter sales, down from 71% a year earlier, according to BigGo Finance.
HRT’s deal follows a similar move by rival Jane Street, which agreed to spend $6 billion on CoreWeave data center capacity and took a $1 billion equity stake in the company, Bloomberg reported.
What the deal means for the AI infrastructure market
Financial firms are becoming increasingly important customers for specialized AI infrastructure providers.
Quantitative trading firms such as HRT and Jane Street have both the budgets and the technical incentive to spend aggressively on computing capacity, particularly as their research teams experiment with larger datasets and more complex machine learning models.
That creates a different kind of opportunity for companies such as CoreWeave. Instead of relying primarily on hyperscalers and AI developers, specialized cloud providers can increasingly look to financial institutions, pharmaceutical companies, research organizations and other compute-intensive industries for large, long-term infrastructure commitments.
The opportunity also extends beyond GPUs. Deployments at this scale require high-speed networking, data center capacity, power, cooling, storage and systems capable of moving enormous volumes of information between customer environments and cloud infrastructure. That broader buildout could create additional demand across the AI infrastructure ecosystem as financial firms expand their computing footprints.
There is still a significant execution challenge. More contracts mean CoreWeave must continue adding capacity quickly enough to support them while managing the debt and capital costs associated with new data centers and hardware. The company reported $2.58 billion in second-quarter revenue, up 112% from a year earlier, and ended the quarter with a $104 billion revenue backlog, according to BigGo Finance.
For now, the HRT agreement adds another major financial-services customer to that pipeline. Combined with Jane Street’s multibillion-dollar commitment, this suggests Wall Street could become a much larger battleground for AI infrastructure providers as trading firms compete for access to the fastest available chips, networks and computing environments.
Related reading: The AI infrastructure race is also spreading higher up the stack, with Stripe agreeing to acquire OpenRouter in a reported $7.5 billion deal that could reshape how businesses manage AI models and spending.