‘Both Hopper and Blackwell systems have certain supply constraints, and the demand for Blackwell is expected to exceed supply for several quarters in fiscal 2026,’ Nvidia finance chief Colette Kress said in her third-quarter commentary before CEO Jensen Huang addressed issues with Blackwell.

Nvidia CFO Colette Kress said the company is “racing to scale supply to meet incredible demand” for its recently launched Blackwell GPUs and associated systems that have been touted to deliver major leaps in generative AI performance and efficiency.

Kress made the comments during the AI computing giant’s Wednesday earnings call, where the company reported that its third-quarter revenue nearly doubled to $35.1 billion mainly thanks to continuing high demand for its data center chips and systems.

[Related: Nvidia Reveals 4-GPU GB200 NVL4 Superchip, Releases H200 NVL Module]

In a prepared commentary on the latest earnings, Kress said Nvidia is scheduled to begin shipping Blackwell products, now in full production, by January, with plans to ramp up those shipments in its 2026 fiscal year, which corresponds with the following months.

“Both Hopper and Blackwell systems have certain supply constraints, and the demand for Blackwell is expected to exceed supply for several quarters in fiscal 2026,” she wrote.

On the call, Kress added that Nvidia is “on track to exceed” its previous Blackwell revenue estimate of several billion dollars for the fourth quarter, which wraps up at the end of January, as its “visibility into supply continues to increase.”

“Blackwell is a customizable AI infrastructure with seven different types of Nvidia built chips, multiple networking options and for air- and liquid-cooled data centers,” she said. “Our current focus is on ramping to strong demand, increasing system availability and providing the optimal mix of configurations to our customers.”

Jensen Huang Addresses Blackwell Issues, Road Map Execution

During the call’s question-and-answer session, a financial analyst asked Nvidia CEO Jensen Huang to address a Sunday report by industry publication The Information, which detailed concerns from a few customers about Blackwell GPUs overheating in the most powerful configurations of its Grace Blackwell GB200 NVL72 rack-scale server platform.

The GB200 NVL72 is expected to serve as the foundation for upcoming flagship Nvidia AI offerings from major OEM and cloud computing partners, including Dell Technologies, Amazon Web Services, Microsoft Azure, Google Cloud and Oracle Cloud.

In response to the question about The Information’s report on Blackwell GPUs overheating in GB200 NVL72 systems, Huang noted that while Blackwell production is in “full steam” with the company exceeding previous revenue estimates, he said the engineering Nvidia does with OEM and cloud computing partners is “rather complicated.”

“The reason for that is because although we build full stack and full infrastructure, we disaggregate all of the AI supercomputer and we integrate it into all of the custom data centers and architectures around the world,” he said on the earnings call.

“That integration process is something we’ve done several generations now. We’ve very good at it, but there’s still a lot of engineering that happens at this point,” Huang added.

Huang pointed out that Dell Technologies announced on Sunday that it had started shipping its GB200 NVL72 server racks to customers, including up-and-coming GPU cloud service provider CoreWeave. He also mentioned Blackwell systems that are being stood up by Oracle Cloud Infrastructure, Microsoft Azure and Google Cloud.

“But as you see from all of the systems that are being stood up, Blackwell is in great shape,” he said. “And as we mentioned earlier, the supply and what we’re planning to shift this quarter is greater than our previous estimates.”

Addressing a question about Nvidia’s ability to execute on its data center road map, which moved to an annual release cadence for GPUs and other chips last year, Huang remained steadfast in his commitment to the accelerated production plan.

“We’re on an annual roadmap, and we’re expecting to continue to execute on that annual roadmap. And by doing so, we increase the performance of our platform. But it’s also really important to realize that when we’re able to increase performance and do so at X factors at a time, we’re reducing the cost of training, we’re reducing the cost of inference, and we’re reducing the cost of AI so that it could be much more accessible,” he said.

Partner Sounds Off On Nvidia Growth, Investor Reaction

After Nvidia released its third-quarter earnings release, investors reacted by sending the company’s stock price down by more than 1 percent in after-hours trading.

While the company beat Wall Street’s expectations on revenue by nearly $2 billion and exceeded the average analyst estimate for earnings per share by 6 cents, its fourth-quarter revenue estimate was only slightly higher than what Wall Street had been anticipating.

Andy Lin, CTO at top Nvidia partner Mark III Systems in Houston, Texas, told CRN that while demand for Nvidia’s data center GPUs and associated systems are “obviously incredibly strong,” it has “set such a high bar” for itself with several quarters of triple-digit growth.

“These numbers are still astounding, especially on the year-over-year compare. But it’s obviously less growth than before,” he said.

However, Lin said, some customers may be holding off from making any purchases right now as Nvidia transitions from Hopper-based systems to Blackwell-based systems.

“There are definitely many organizations we see that have definitely not spent [money on new infrastructure and are waiting] on Blackwell. So the question is, how many of those there are, to what scale and what that’s going to look like? And I think that could be one thing that maybe the market is underestimating a little bit,” he said.