Corvex, Inc.는 약 1.5의 중요한 IT 전력 용량을 확장하기 위해 3,300만 달러 규모의 자금 조달 라운드를 발표했습니다.

Corvex, Inc., an engineering-led AI computing platform specializing in secure, GPU-accelerated infrastructure, announced plans to increase its critical IT power capacity from approximately 1.5 MW to approximately 8 MW by the end of 2026. The expansion involves doubling capacity at its existing Mid-Atlantic data center and launching operations at a second enterprise-scale facility in the Midwest. Both sites are scheduled to be operational in the fourth quarter of 2026, with revenue generation expected to begin in the first quarter of 2027. The deployments will bring approximately 3,000 latest-generation GPUs online, split between roughly 2,000 at the Midwest site and 1,000 at the Mid-Atlantic site. These clusters will be offered as bare metal or with a managed Kubernetes orchestration layer. Additionally, Corvex holds a right of first refusal on an extra 12.5 MW of capacity at the Midwest location, which could raise total critical IT power to over 20 MW if exercised, representing a 13-fold increase from current levels. That additional capacity is targeted for service in the third quarter of 2027, pending capital availability, customer commitments, equipment delivery, utility energization, and local permitting.

To finance the expansion, Corvex secured a 33 million dollar private placement of common stock and Series D Preferred Stock, led by Goldman Sachs & Co. LLC, Morgan Stanley, and Oppenheimer & Co. as Joint Lead Placement Agents. The company entered into a securities purchase agreement to issue 4.258 million shares of restricted common stock at 7.75 dollars per share, with closing anticipated around September 2, 2026. The transaction is projected to boost pro forma cash and cash equivalents to approximately 55 million dollars, up from 22 million dollars reported as of June 30, 2026. Proceeds will fund the capacity expansion and advance development of the Amplified AI Cloud platform, Token Factory, and Assured AI offerings. A portion of the funds will accelerate the Token Factory, currently in closed alpha, which provides API access to high-performing open-source models with a zero data retention policy on a SOC 2 Type II certified platform supporting HIPAA compliance. Funds will also support Assured AI, a zero-trust secure computing platform designed to protect privacy-sensitive inference requests and customer model weights, which will be integrated into the Token Factory.

Corvex has executed definitive agreements for both expansions, leveraging existing enterprise data centers with live utility power to reduce development risk and shorten deployment timelines compared to greenfield construction. The company is actively discussing capacity utilization with potential customers, aiming to secure multi-year, take-or-pay commitments for long-term GPU-as-a-Service contracts at prevailing market rates, alongside high-throughput storage and CPU compute services. Jay Crystal, co-CEO and co-founder of Corvex, stated, The scarce input in AI infrastructure is energized power inside a finished building. We have secured more than five times as much of it, with the right to scale to more than 20 megawatts in 2027. Seth Demsey, co-CEO and co-founder, added, As inference volumes grow, enterprises are paying much closer attention to the cost of serving AI at scale. Token Factory is built to address that cost by offering near-frontier performance on most enterprise workloads at a fraction of the price per million tokens of frontier proprietary models. The differentiation will be in the infrastructure and security layer beneath the API. Integrating our Assured AI security capabilities directly into the inference platform gives customers a way to combine lower-cost inference with greater control over sensitive prompts and model weights. The securities issued in the private placement are unregistered under the Securities Act of 1933 and are being sold under exemptions, with Corvex agreeing to file a registration statement for their resale. Standard forward-looking statement disclaimers apply regarding future performance and customer agreements.