NVIDIA GPU Rental Prices Double as Memory Supply Commitments Reach $390 Billion

NVIDIA's AI computing business continues expanding at an extraordinary pace, with interim CEO Colette Kress revealing that some older Hopper GPUs are being rented for up to twice their original contract prices. Meanwhile, the company has increased its supply commitments to $390 billion, primarily driven by multiyear memory procurement agreements as demand for advanced AI hardware continues exceeding available production capacity.

During NVIDIA's fiscal 2027 third quarter earnings presentation on November 17, Kress explained that declining AI inference costs are accelerating demand for computing resources. Token processing prices have fallen nearly 60% since May, encouraging greater AI adoption and increased GPU utilization across cloud providers. Consequently, some Hopper accelerators have secured renewed multiyear rental contracts at up to twice their original rates, while Blackwell Ultra B300 rental prices have increased more than 60% since their introduction.

"GPU rental prices are rising across generations."
— Quote by: Colette Kress, NVIDIA Interim CEO.

The trend reinforces the growing rental premiums for NVIDIA's older AI accelerators, demonstrating that previous generation hardware remains valuable for inference and other computing workloads despite newer architectures entering the market. However, the reported price increases refer to specific cloud computing agreements rather than universal rental rates across every provider.

According to NVIDIA's financial disclosures filed with the SEC, the company's supply and capacity commitments increased from $279 billion to $390 billion, largely reflecting the cost of securing memory over multiple years. NVIDIA also disclosed $102 billion in cloud service agreements and future data center lease commitments, highlighting the scale of its infrastructure investments.

The aggressive procurement strategy comes as the global memory industry faces continued supply constraints, with SK Hynix warning that shortages could persist through 2030. NVIDIA is expanding agreements with manufacturing and packaging partners to secure additional capacity for its Blackwell and Vera Rubin platforms.

Despite increasing component costs, NVIDIA reported $115.8 billion in quarterly revenue, representing 103% annual growth, with its data center business contributing $107.6 billion. The company expects fourth quarter revenue of approximately $130 billion, demonstrating continued confidence in AI infrastructure demand.

However, NVIDIA also acknowledged that higher memory and system prices are affecting consumer PC demand, raising concerns about how enterprise AI investments could influence the affordability and availability of gaming hardware.

NVIDIA's financial results demonstrate that demand for AI computing remains strong, even as the industry faces increasingly expensive memory and infrastructure requirements. However, committing $390 billion to future supply also illustrates how aggressively major corporations are securing manufacturing capacity.

For PC gamers and local AI enthusiasts, this concentration of resources is concerning. Memory manufacturers increasingly prioritize profitable enterprise contracts while consumer components face supply restrictions and higher prices.

Although enterprise computing represents a lucrative market, it should not come at the expense of accessible gaming hardware, affordable memory, and independent PC ownership. NVIDIA and its suppliers must maintain a sustainable balance between expanding AI infrastructure and supporting the consumer market.

With NVIDIA committing $390 billion to future supply and AI GPU rental prices continuing to rise, do you believe consumer graphics cards and memory will become even more expensive?

Share
Angel Morales

Founder and lead writer at Duck-IT Tech News, and dedicated to delivering the latest news, reviews, and insights in the world of technology, gaming, and AI. With experience in the tech and business sectors, combining a deep passion for technology with a talent for clear and engaging writing

Previous
Previous

Zelda: Majora's Mask Remake Rumors Grow, but Insiders Say Development Has Not Started

Next
Next

EXODUS Reveals Female Jun, Confirms Voice Cast and Official Artbook Ahead of 2027 Launch