Nvidia is securing more components than ever to keep pace with booming demand for artificial intelligence infrastructure. But even with a major increase in its supply commitments, the company says shortages across its supply chain are limiting how quickly its AI-related revenues can grow.
During its Aug. 26 earnings call, Nvidia executives warned that supply constraints are weighing on the company’s financial outlook for its next fiscal year. The problem is not limited to a single component or supplier. According to co-founder and CEO Jen-Hsun Huang, shortages stretch across the company’s vast network, from high-speed memory and silicon suppliers to companies involved in building AI data centers.
“We have a really gigantic supply chain,” Huang told investors, pointing to the scale of Nvidia’s supplier network. The company has secured significant volumes, he said, but still needs considerably more to satisfy market demand.
Nvidia currently has enough supply to support 70% year-over-year revenue growth in fiscal 2028, according to Huang. That would represent substantial expansion, but it remains below the level of demand the company is seeing.
AI infrastructure boom puts pressure on semiconductor supply
The supply squeeze reflects a broader problem across the technology industry. Demand for the infrastructure needed to power artificial intelligence data centers has grown faster than semiconductor supply, with high-speed memory emerging as one of the most significant constraints.
Other technology companies, including Dell Technologies and Hewlett Packard Enterprise, have been competing with suppliers for additional semiconductor capacity. The shortage has also contributed to longer waiting times for enterprise servers designed to run AI workloads.
Nvidia is directly exposed to that pressure because its customer base includes organizations building large-scale AI data centers. Among them are major companies such as Amazon, Google and Microsoft, as well as national and regional governments.
For Nvidia, the imbalance between supply and demand is unlikely to disappear quickly. Executive Vice President and Chief Financial Officer Colette Kress said the bottleneck is expected to persist at least through the fiscal year ending in early 2028.
Memory prices are rising sharply
The supply shortage is also affecting component pricing. Kress said semiconductor shortages, particularly in memory, have pushed prices higher, with the market currently facing what she described as “extreme pricing conditions in memory.”
“The magnitude of the price increase has exceeded our prior expectations and are headed even higher into next year,” Kress said.
Nvidia does, however, have an advantage in its relationships with suppliers. The company has long-standing partnerships with three major memory suppliers and is working closely with them to expand production capacity, according to Kress.
The company has also moved to secure critical components well beyond the immediate production cycle. According to a securities filing, Nvidia has entered into supply commitments covering the coming years.
Those commitments increased dramatically during the second quarter. They rose from $119 billion in the first quarter to $279 billion as of the end of the period ended July 26 — an increase of $160 billion in a single quarter.
Nvidia is looking beyond its own component needs
The scale of the shortage is forcing Nvidia to take a broader approach to supply-chain planning.
Rather than simply securing the components needed to manufacture its products, the company is increasingly trying to understand the supply networks supporting the construction of AI data centers themselves.
Huang said Nvidia is aligning its supply chain with the broader networks feeding data-center construction. The goal is to gain greater visibility into when customers will need equipment and infrastructure in the future.
The reason is straightforward: Nvidia’s customers are no longer simply ordering individual products. They are building entire AI infrastructure platforms, often involving multiple layers of equipment, components and facilities.
“They really need an entire factory platform built for them,” Huang said, referring to the needs of data-center customers.
For Nvidia, the challenge is therefore no longer simply a question of producing enough chips. The company must coordinate a sprawling supply ecosystem capable of supporting the extraordinary pace of AI infrastructure construction — at a time when demand continues to outstrip available capacity.
