Nvidia CEO Jensen Huang pushes back on claims of Vera Rubin manufacturing delays, says next AI accelerator is already in production

Nvidia’s Jensen Huang has denied reports that the company’s next-generation Vera Rubin AI accelerator platform is facing manufacturing setbacks. In comments relayed by Yahoo Finance, the Nvidia CEO said production for the platform is already underway, directly disputing claims that the launch schedule could slip.

The remarks come as investors have remained focused on the cadence of Nvidia’s data center product cycle, especially for AI compute systems that power training and inference for large language models and other workloads. Nvidia has positioned its accelerator roadmap around successive generations of chips designed to increase performance-per-watt and improve overall system efficiency, which is a key determinant of demand from cloud and enterprise customers.

In the report, Huang was responding to the narrative that Vera Rubin could be delayed. He characterized that concern as inaccurate, saying the platform is in production rather than being held up. The comments were presented as a denial of any manufacturing-driven timing problem rather than as a new forecast of when customers will fully receive systems at scale.

The specific source coverage did not detail the nature of any earlier delay claims, nor did it offer granular supply-chain information. It also did not disclose whether Nvidia has already ramped at a particular factory utilization rate, whether output is constrained by packaging, memory supply, or substrate availability, or how any potential capacity limitations might be mitigated.

Vera Rubin is understood in Nvidia’s broader AI strategy as a forthcoming accelerator platform, meaning it is intended to succeed current architectures used across the company’s data center lineup. For Nvidia, these platforms matter because they drive both direct GPU revenue and the broader “system” ecosystem around networking, software stacks, and developer tooling that make it easier for customers to deploy AI at scale.

Market participants often look for confirmation that new accelerator platforms are progressing smoothly, since any manufacturing disruption can cascade into customer delivery timelines, partner system availability, and expectations for future revenue growth. By saying the platform is already in production, Nvidia is indicating that the near-term risk of a schedule slip tied to manufacturing is lower than some coverage had suggested.

Still, the company did not provide additional timing specificity in the disclosed remarks, such as when Nvidia expects peak shipment volumes, when the first customer systems are broadly available, or whether particular product configurations will arrive in a staged rollout. Without those details, it remains unclear how quickly the company will translate “in production” status into measurable shipments at the pace the market may be expecting.