NVIDIA announces its next-gen B100 AI GPU will be supply constrained as demand far exceeds supply.

NVIDIA posted its fourth fiscal quarter earnings, surpassing Wall Street predictions for earnings and sales. The company reported record revenue for the quarter, driven by its Data Center business, and provided a preview of its next-generation Blackwell B100 AI GPU. CEO Jensen Huang explained the challenge of scaling new products, noting: "Whenever we have new products, as you know, it ramps from zero to a very large number and you can't do that overnight." The upcoming products driving this momentum are NVIDIA's beefed-up H200 AI GPU and its next-generation B100 AI GPU, both expected to launch this year.

Although NVIDIA has recently improved its supply chain issues that previously hampered AI GPU shipments, the company continues to sell AI GPUs faster than it can manufacture them. Chief Financial Officer Colette Kress confirmed this dynamic, stating: "We are delighted that supply of Hopper architecture products is improving. Demand for Hopper remains very strong. We can expect our next-generation products to be supply constrained as demand far exceeds supply." The supply constraints stem from NVIDIA's reliance on TSMC (Taiwan Semiconductor Manufacturing Company), which manufactures NVIDIA's entire fleet of AI GPUs including the current H100, new H200, and next-generation B100 models. The advanced CoWoS packaging technology required for these chips limits production capacity.

NVIDIA's next-generation Blackwell GPU architecture is scheduled to be unveiled at the company's GPU Technology Conference (GTC), taking place on March 18. CEO Jensen Huang will deliver a keynote address, where the company plans to provide full unveiling and announcement of the B100 AI GPU along with benchmarks and detailed specifications. Given the already exceptional demand for the H100 and H200 models, the B100 is expected to face significant supply constraints as demand far exceeds what TSMC can produce.