NVIDIA’s Blackwell Ultra GB300 AI servers are drawing extraordinary demand, with supplier Quanta warning that current production can’t keep pace with orders. The company expects shipments of NVIDIA AI systems to crest by the fourth quarter, and says demand should remain robust for years as enterprises and cloud providers race to scale their AI infrastructure.
Quanta, a key manufacturing partner alongside other major suppliers, describes interest in NVIDIA’s GB200 and GB300 platforms as far beyond initial projections. The surge in AI spending has created a supply crunch that isn’t easing yet, and Quanta believes the momentum will carry well into the future rather than being a short-lived spike.
To relieve pressure and add capacity closer to customers, Quanta is expanding manufacturing beyond its existing bases. New production lines are being set up in the United States, Thailand, and Mexico. The Mexico facility is slated to begin output next year, with additional investments underway in California and Tennessee to support the build-out of AI server capacity. By year-end, Quanta expects a shipment peak heavily weighted toward Blackwell Ultra GB300 clusters.
NVIDIA’s Blackwell Ultra lineup did encounter a brief speed bump tied to a board redesign. The company shifted from the Bianca board to the modular Cordelia design featuring SOCAMM memory, but the abrupt change strained the supply chain. To keep momentum, NVIDIA pivoted back to a GB200-style board, enabling a smoother and faster production ramp while suppliers adjusted.
With that pivot complete, production has accelerated and orders continue to pour in. Supply chain chatter points to sustained strength through the end of the year, as enterprises standardize on NVIDIA’s GB200 and GB300 systems for training and inference across increasingly large AI models.
Adding to the tailwind, the upcoming Vera Rubin platform is expected to debut soon, positioning NVIDIA for another lift in the AI data center market. If current trends hold, Blackwell Ultra GB300 servers could define the next wave of AI infrastructure growth, with Q4 shaping up to be a pivotal quarter for shipments and deployments.
