The move highlights the ongoing shortage of advanced AI chips, forcing even Nvidia to rent back its own hardware to secure access for cloud operations
In sum – what to know:
$1.5B lease with Lambda – Nvidia signs major GPU rental deal for 18,000 servers, underscoring demand pressures and supply constraints in AI compute.
IPO ambitions for Lambda – The deal makes Nvidia its largest customer, boosting the company’s profile as it prepares to go public.
Chip shortage drives strategy – Even Nvidia is leasing back its own GPUs to secure capacity, while supporting smaller providers to challenge hyperscale cloud rivals.
Nvidia has agreed to a $1.5 billion contract to lease GPU servers from Lambda, an AI infrastructure company offering on-demand GPUs, according to a report by The Information.
The agreement covers a $1.3 billion, four-year lease for 10,000 servers, plus a $200 million contract for 8,000 additional servers. Nvidia, already an investor in Lambda, becomes its largest customer under the arrangement, according to the report.
The move highlights the ongoing shortage of advanced AI chips, forcing even Nvidia to rent back its own hardware to secure access for cloud operations. The deal also strengthens smaller providers competing with hyperscale players Amazon, Microsoft, and Google.
Lambda projects its cloud revenues could surpass $1 billion by 2026, positioning itself as a significant alternative for AI developers seeking access to GPUs. The AI chip giant has previously backed similar companies, including CoreWeave, as part of a broader strategy to diversify AI compute supply.
Lambda has reportedly hired Morgan Stanley, J.P. Morgan, and Citi for a public listing that could happen as early as the first half of 2026.
Nearly 40% of Nvidia’s revenue in the second quarter of the year came from just two customers, according to a filing with the U.S. Securities and Exchange Commission.
The chipmaker recently announced record revenue of $46.7 billion for the quarter ending July 27, a 56% increase from the previous year, fueled by surging demand for AI data center hardware.
The filing shows that one customer represented 23% of Nvidia’s Q2 revenue, while another accounted for 16%. The company did not name them, referring only to “Customer A” and “Customer B.
For the first half of its fiscal year, the firm reported that these two customers contributed 20% and 15% of total revenue, respectively. Four additional customers made up 14%, 11%, 11%, and 10% of revenue in the second quarter.
