Huang says Vera Rubin in production, dismisses delay reports

Investing.com -- NVIDIA (NASDAQ: NVDA) CEO Jensen Huang flatly denied reports that the company's next-generation Vera Rubin AI accelerator system is facing manufacturing setbacks, declaring "Vera Rubin is already in production. Giant amounts of production incoming" at a developer event in Tokyo on Wednesday, according to Bloomberg. The pushback targets a claim from research firm SemiAnalysis, which reported this month that the Vera Rubin AI server rack system had been delayed due to difficulties manufacturing a specialized circuit board that connects electronic modules.

The denial of a delay is notable. Any credible delay in Vera Rubin's ramp would sharpen competitive pressure from AMD's advancing MI350 and MI400 roadmap, making the CEO's on-record rebuttal particularly consequential for investors who have priced Nvidia's product cycle into its elevated valuation.

Bloomberg reported that Huang made the remarks to reporters on the sidelines of the Tokyo developer event, delivering a one-word dismissal of the SemiAnalysis findings before elaborating on production volumes. SemiAnalysis has built a reputation for detailed Nvidia supply chain analysis, making its delay report notable enough to warrant a direct response from the chief executive. The specific technical challenge cited — manufacturing a circuit board that bridges electronic modules within an AI server rack — points to the kind of packaging complexity that has historically been a bottleneck for next-generation accelerator systems.

The Tokyo setting was not incidental. Japan's Prime Minister Sanae Takaichi has unveiled an unprecedented $2.3 trillion spending plan aimed at encouraging private sector investment in domestic AI, and SoftBank Group has separately announced a $5.4 billion deal to acquire ABB Ltd.'s robotics unit, according to Bloomberg. Japan's shrinking population has long pressured manufacturers to accelerate adoption of robotics and factory automation, with major players including Fanuc Corp., Yaskawa Electric Corp., Kawasaki Heavy Industries, and Toyota Motor Corp. all active in the space. For Nvidia, whose physical AI and robotics platforms are increasingly central to its pitch beyond the data center, Japan represents a strategically important market.

Huang also addressed the status of Nvidia's H200 chip shipments to China, an issue that has drawn significant investor attention since Washington eased export restrictions. The H200 was cleared for export by President Donald Trump in December, Bloomberg reported, marking a notable relaxation of the controls designed to limit Beijing's access to advanced AI hardware. Until Blackwell-architecture chips arrived in late 2024, the H200 was the most powerful AI accelerator on the market, and its availability to Chinese buyers carries meaningful revenue implications for Nvidia.

Despite the green light from Washington, Huang indicated the ramp has barely begun. "We haven't really started much shipment yet," he told reporters, as Bloomberg noted. A Trump administration official overseeing export control programs confirmed that a small number of H200 chips have already reached Chinese customers, according to Bloomberg, suggesting the logistics and licensing framework is in place even if volumes remain modest.

The combination of a Vera Rubin production affirmation and an H200 China ramp still in its early stages gives investors two distinct forward-looking signals to monitor. On Vera Rubin, the absence of any specific volume or delivery timeline from Huang himself means the SemiAnalysis debate will likely persist until major cloud customers or hyperscalers publicly confirm order receipts or deployment schedules. Nvidia has not provided public guidance on Vera Rubin delivery windows, and no customer order commitments have been disclosed.

On China, the H200 shipment trajectory will be closely watched as a potential revenue catalyst. China has historically represented a significant portion of Nvidia's data center revenue before export restrictions curtailed sales, and even a partial restoration of that demand could provide a meaningful uplift to results in the coming quarters. How quickly Nvidia can scale H200 deliveries to Chinese buyers, and whether additional license conditions constrain volumes, will be key questions heading into Nvidia's next earnings report.

For now, Huang's unambiguous language in Tokyo offers the clearest possible executive-level signal that Nvidia views Vera Rubin's timeline as intact, setting up a direct credibility test between the CEO and one of the semiconductor industry's most closely followed independent research firms.

You May Also Be Interested In

• Intel mulls partnership for Ohio chip fab, SK Hynix among candidates - report
• Active options: NVDA AAPL SMCI MU AMZN TSLA MSFT INTC NFLX PLTR
• White House accuses Moonshot of violating US export controls using Nvidia hardware