THE APEX TIMES
Hudson River Trading signs multi-year deal to run quantitative research on CoreWeave cloud using NVIDIA’s Vera Rubin platform
The quantitative trading firm is partnering with CoreWeave under a multi-year agreement, using NVIDIA’s Vera Rubin NVL72 infrastructure on the cloud to support next-generation research computing.
Hudson River Trading (HRT) said it will build a next-generation research platform with computing supplied through CoreWeave, a cloud provider focused on artificial intelligence workloads. The company described the effort as a multi-year agreement announced on August 20, 2026, positioning the project as a way to accelerate research and development for quantitative trading strategies.
According to the announcement carried by Yahoo Finance, HRT will base the platform on NVIDIA’s Vera Rubin infrastructure, specifically referencing the Vera Rubin NVL72 configuration. NVIDIA’s Vera Rubin is designed for large-scale AI training and accelerated data processing, and the NVL72 label points to a particular system configuration within that platform family.
CoreWeave is described in the same report as supplying the cloud environment for the build. In practical terms, the move reflects a broader shift in advanced markets research, where firms increasingly treat high-performance compute as an on-demand capability rather than relying exclusively on in-house hardware.
For HRT, the primary business objective is likely speed and throughput for experimentation. Quantitative trading teams typically run repeated simulations, model training, and feature evaluation on large volumes of market and alternative data. Larger or faster compute resources can shorten the cycle between hypothesis and test, which can matter when strategies are sensitive to regime changes and execution constraints.
The deal also underscores how AI infrastructure is moving from experimentation into the core tooling of financial research. NVIDIA has been promoting data center and AI systems that can scale across training and inference, while cloud providers like CoreWeave target customers that need substantial acceleration without building and operating their own specialized clusters.
NVIDIA’s Vera Rubin platform, as referenced here, is part of that push to standardize high-performance systems for accelerated workloads. While the specific performance targets and deployment details for HRT are not included in the report, the involvement of a named configuration suggests HRT is seeking a tightly defined compute footprint rather than a general-purpose setup.
The announcement does not disclose the financial terms of the agreement, the expected timelines for deploying the platform, or the exact types of workloads HRT plans to prioritize first. It also does not say whether the arrangement includes additional services beyond compute access, such as managed infrastructure support or software integration.
What to watch next is whether HRT provides more detail in subsequent communications about the platform’s capabilities, such as support for new model training pipelines, data ingestion approaches, or latency-sensitive research processes. Investors and industry observers will also look for any follow-on references from CoreWeave or NVIDIA that clarify how the Vera Rubin NVL72 configuration is being applied in this financial context.
Why It Matters
- Cloud-based accelerated computing is increasingly becoming a building block for quantitative research, where compute speed can affect model development cycles.
- Named use of NVIDIA’s Vera Rubin configuration suggests a move toward more standardized, performance-oriented AI infrastructure rather than ad hoc resource procurement.
- The deal reflects continued investment by advanced market participants in AI-adjacent compute stacks, even when the underlying trading strategies remain proprietary.
- For CoreWeave and NVIDIA, the partnership is another announcement of demand for high-performance AI infrastructure beyond traditional technology sectors.
Key Facts
- HRT announced a multi-year agreement with CoreWeave to build a next-generation research platform.
- The platform is described as being powered by NVIDIA’s Vera Rubin infrastructure, with the report specifically citing the Vera Rubin NVL72 configuration.
- CoreWeave is positioned as the provider of the cloud environment for the project.
- The announcement was published on August 20, 2026, as carried by Yahoo Finance.
- NVIDIA’s involvement centers on using its Vera Rubin platform for accelerated computing in support of research workloads.
Technology Related
CoreWeave signs multibillion-dollar Hudson River Trading deal, underscoring Wall Street’s scramble for Nvidia’s latest AI chips
CoreWeave Inc. has reached a multiyear agreement with Hudson River Trading for access to Nvidia Corp. chips, according to a report. The move highlights how trading firms and data-center players are competing for constrained compute capacity tied to the AI chip cycle.
Apple shares face a make-or-break window as investors watch a cluster of product catalysts
A fresh market outlook argues Apple is approaching a key stretch in its product roadmap, hinging on a redesigned iPhone lineup, the rollout momentum of new Apple silicon, and an anticipated upgrade to Siri powered by artificial intelligence.
Wedbush flags limited near-term impact from NVIDIA’s China H200 return
Even if NVIDIA resumes meaningful shipments of its H200 data center AI GPU into China, a key Wall Street view suggests revenue upside could be modest in the short run.
Broadcom shares hold steady as Google’s Marvell custom-chip deal reshuffles the chip-news tape
A $12.2 billion equity-linked custom-chip award to Marvell put AI-related semiconductor expectations back in focus, leaving Broadcom investors watching how the latest customer moves could affect the broader AI supply chain.
Walmart’s Miss Sparks a Broader Anxiety Over Consumer Demand, Pulling Nvidia Into Focus
A rare Walmart sales miss and a sharp one-day selloff reignited trader worries that the US consumer may be losing momentum, adding pressure to high-expectations stocks tied to the technology spending cycle.
YouTube tells creators to steer clear of simultaneous Netflix postings, warning of lost marketing support and brand revenue shares
In a new warning to content creators, YouTube said publishing the same video on Netflix at the same time as YouTube can reduce promotional support and affect how creators share revenue from brand campaigns.
NVIDIA adds Firefox browser support to GeForce NOW, aiming to make cloud PC gaming easier to access
GeForce NOW, NVIDIA’s cloud gaming service, now runs in Firefox on Windows, reducing the need for users to install separate apps and expanding where supported PC games can be played from the browser.
Meta faces new legal exposure as child-targeting trial raises fresh concerns about risk and spend
A trial alleging Meta’s social platforms engaged in child-targeting behavior adds a potentially massive liability overhang for the company, coming as its stock remains under pressure from investors’ questions about artificial-intelligence spending.
Intel research finds executives want robot fleets soon, but fewer have plans to scale them
A new Intel study estimates that 6 in 10 senior leaders expect their organizations to run robot fleets within five years, yet only 4 in 10 have a formal strategy for a mixed human-robot workforce.
Tiger Global investor Chase Coleman reportedly trimmed Nvidia and added to other AI-chip bets
A Yahoo Finance report says Chase Coleman of Tiger Global reduced his Nvidia position in the second quarter while building positions in Cerebras and Intel, underscoring how quickly AI-chip portfolios are rotating.