NVIDIA has provided substantial support to Lambda, an artificial intelligence cloud service provider preparing for an initial public offering (IPO), according to industry sources.

Sources revealed that NVIDIA agreed this summer to lease 10,000 GPU servers equipped with NVIDIA's AI chips from Lambda over a four-year period, valued at $1.3 billion. Additionally, NVIDIA reached another $200 million agreement to rent 8,000 servers equipped with NVIDIA chips, though the timeline remains undetermined. These contracts make NVIDIA Lambda's largest customer to date and establish a foundation for the company's upcoming IPO.

This represents the latest example of NVIDIA's "circular" financial arrangements to drive its chips into the cloud market and help smaller cloud service providers compete against traditional giants like Amazon and Google. It also demonstrates the "internal circulation" of capital in the AI sector, where NVIDIA simultaneously serves as supplier, investor, and customer, supporting multiple "neocloud" companies.

**Business Model Similar to CoreWeave**

Lambda's business model involves leasing data center space, deploying servers equipped with NVIDIA GPUs, and contracting with customers to rent these servers. The specific costs Lambda charges NVIDIA for GPU server rentals and how NVIDIA accounts for these transactions financially remain unclear, given NVIDIA's role as both buyer and seller, as well as a Lambda shareholder.

Another source indicated that NVIDIA's own researchers will also utilize the GPU servers leased from Lambda.

Beyond NVIDIA, Lambda's other major clients include Amazon and Microsoft, which together contributed nearly half of the company's approximately $114 million in cloud revenue during the second quarter. Notably, Amazon and Microsoft primarily use Lambda's GPU servers for internal purposes rather than for AWS or Azure platform customer services.

Lambda projects its cloud revenue will exceed $1 billion by 2026 and $20 billion by 2030, targeting contracts with major AI developers including OpenAI, Google, Anthropic, and xAI.

Lambda also expects its computing capacity to reach nearly 3 gigawatts by 2030, equivalent to nearly half the total computing power of some of the largest current cloud service providers, up from just 47 megawatts in the second quarter. How the company will achieve this growth remains unclear, though going public could help expand operations through debt financing.

Lambda's business model and high customer concentration resemble CoreWeave, a larger GPU cloud service provider that recently went public and similarly received substantial NVIDIA support.

Previously, Lambda primarily signed smaller-scale, short-term GPU rental contracts. This deal with NVIDIA represents its largest transaction to date and will likely support marketing efforts ahead of its anticipated IPO in the first half of next year.

**Supporting Smaller Companies as NVIDIA's Consistent Strategy**

NVIDIA consistently supports companies willing to use its chips and more receptive to purchasing diverse hardware products compared to traditional cloud giants. For instance, NVIDIA has clashed with Microsoft over GPU server rack designs, while Lambda executives have internally discussed adopting NVIDIA's developing optical networking technology.

NVIDIA previously helped CoreWeave rise rapidly. The cryptocurrency mining-turned-cloud company signed nearly identical agreements with NVIDIA early in its transformation. These agreements helped CoreWeave secure debt financing and expand its cloud business, capturing market share from traditional cloud providers.

NVIDIA's support for smaller cloud service providers aims to protect its core business long-term. While NVIDIA's largest customers remain Microsoft, Amazon, and Google, these tech giants are also developing proprietary AI chips to reduce dependence on NVIDIA.

**Customer Concentration Risk**

Although Amazon and Microsoft purchase far more GPUs for their own data centers than they lease from third parties like Lambda, both companies report their GPU servers operate at near-full capacity around the clock, with data center expansion unable to keep pace with demand.

Microsoft's contract with Lambda is much smaller than its leasing agreement with CoreWeave, but Lambda executives indicate the company is negotiating larger-scale partnerships with other potential clients.

However, Lambda's ability to secure such deals remains uncertain. Company leadership acknowledges that, like other cloud service providers, Lambda faces power supply and data center space constraints.

CoreWeave previously relied on external data center leasing but recently acquired a major power and data center company for $9 billion, planning to build proprietary sites to reduce costs.

**NVIDIA Partnership Details**

Lambda executives stated that NVIDIA's $1.3 billion GPU leasing agreement, codenamed "Project Comet," will support its emerging DGX Cloud computing business. Through this platform, NVIDIA leases GPUs from cloud service providers and subleases them to AI-developing enterprises, while NVIDIA's own researchers also use the platform.

Analysis suggests NVIDIA values Lambda for multiple reasons, including Lambda's success in attracting more customers to switch to NVIDIA GPUs. For example, Lambda recently signed a one-year partnership with image generation startup Midjourney, helping the company migrate code originally running on Google AI chips to NVIDIA's next-generation Blackwell GPUs.

Lambda executives noted that converting Google AI chip users to NVIDIA GPU users has elevated the company's standing within NVIDIA.

Google's Tensor Processing Unit (TPU) chips have become increasingly competitive in AI in recent years. Google has also approached GPU-focused cloud service providers like CoreWeave, hoping they would deploy Google chips, with at least one company agreeing to cooperate.