AI Computing Power Capital Battle Upgrades Again! Nvidia-backed "New Cloud" Lambda Plans to Raise $3 Billion to Fuel Its Computing Power Landscape and Prepare for IPO

According to Zhitong Finance APP, Lambda, a new AI cloud computing service provider supported by Nvidia, also known as “Neocloud,” is considering raising up to $3 billion in a pre-IPO round, targeting a valuation as high as $12 billion or more, and has already received multiple term sheets from investors. If the deal goes through, this will become a key capital injection round for Lambda ahead of its potential IPO next year, marking a considerable test of private financial markets’ demand, growth visibility, and valuation ranges for new cloud computing platforms.

Lambda is expected to surpass $1.5 billion in revenue this year, and the intended fundraising could be about twice its annual revenue, highlighting the AI compute business’s heavy reliance on capital for GPU procurement, power capacity, and data center construction. Nvidia’s shareholder position strengthens Lambda’s compute supply and ecosystem endorsement, but the ongoing infrastructure and compute rental expansion race among Nebius, Nscale, CoreWeave, and IREN also means investors will ultimately focus on Lambda’s GPU utilization rate, customer concentration, capital efficiency, and ability to generate free cash flow.

Neocloud Lambda considering raising up to $3 billion ahead of IPO

According to reports citing sources, AI cloud computing service provider Lambda, backed by “AI chip superpower” Nvidia (NVDA.US), is negotiating a funding round of up to $3 billion in preparation for a potential IPO next year.

The report adds that this new cloud service provider is discussing raising capital at a valuation as high as $12 billion or more. Negotiations are still ongoing, and deal terms have yet to be finalized.

The article notes that Lambda has received several term sheets for this round of funding, and some people close to the company suggest this round may pave the way for its IPO as early as next year.

According to the company, the California-based enterprise expects to achieve over $1.5 billion in revenue this year.

Lambda did not immediately respond to any media requests for comment.

In November last year, Lambda raised more than $1.5 billion in a round led by TWG Global. Other backers included Andra Capital, SGW—Scott Hassan’s family office known for early investments in Google, OpenAI co-founder Andrej Karpathy, Ark Invest led by top Wall Street fund manager “Cathy Wood,” and Nvidia’s own venture arm, among others.

The company faces fierce competition from other new cloud computing service providers, such as Nebius (NBIS.US), Nscale, CoreWeave (CRWV.US), and IREN (IREN.US). According to reports, London-based Nscale is seeking to raise up to $3 billion through an IPO in the US.

GPU leasing enters center stage in capital markets! “Neocloud”-type AI compute factories enter the IPO era

Founded in 2012, Lambda is a pure AI infrastructure company—not a large model developer or traditional data center real estate provider. Its main business involves building and operating “AI factories” centered on Nvidia GPUs, leasing compute for training, fine-tuning, and inference via the cloud. Its product suite covers on-demand instances with 1–8 Nvidia AI GPUs, one-click compute clusters of 16–2,000+ AI GPU chips, and dedicated AI superclusters with 4,000–165,000+ GPUs on contracts longer than three years, integrating high-density power, liquid cooling, high-speed optical interconnects, and network infrastructure cluster operations.

Lambda has ceased its traditional on-premises workstation and server leasing business as of 2025, fully pivoting to AI cloud computing and hyperscale dedicated AI compute infrastructure delivery; its multi-year agreement with Microsoft involves deploying tens of thousands of Nvidia GPUs.

Lambda and CoreWeave are both “Nvidia-preferred” new cloud service providers (Neocloud), but there are significant differences in platform maturity and depth: Lambda is more focused on GPU compute, dedicated superclusters, and co-engineering services with a relatively simple product structure; CoreWeave has evolved into a full-stack AI cloud platform covering bare-metal Kubernetes, Slurm scheduling, object and distributed storage, high-speed networking, dedicated and serverless inference, and AI agent sandboxes.

As of the end of March 2026, CoreWeave owns 49 data centers, over 1 GW of operating capacity, and more than 3.5 GW of contracted capacity. First-quarter revenue reached $2.078 billion, with a revenue backlog approaching $100 billion. In comparison, Lambda expects annual revenue exceeding $1.5 billion—a clearly smaller scale—but its dedicated clusters and investment-grade customer contracts with clients like Microsoft position it closer to a “focused physical cluster-type AI compute factory operator.”

Lambda's plan to raise up to $3 billion at a $12 billion+ valuation highlights how AI compute infrastructure resources are now forming a full capital cycle of “long-term customer contracts—secured loans and project financing—equity financing—IPO.” The company has previously received a $1 billion syndicated secured credit line and completed $926 million in Moody’s Baa2-rated term loan financing. At the industry chain level, this means Nvidia GPUs, HBM, NVLink/InfiniBand interconnects, optical communication, liquid cooling, and data center power still command strong order visibility.

However, a $12 billion valuation is roughly 8 times its expected annual revenue. Going forward, capital markets will no longer reward GPU volume alone but will scrutinize compute utilization, quality of long-term contracts, customer concentration, GPU depreciation and upgrade cycles, financing costs, and free cash flow. This fundraising proves AI infrastructure can still attract massive capital, but does not in itself prove that downstream AI monetization or the Neocloud profit model has entered a phase of mature growth.