Nvidia CEO Huang denies rumors of Vera Rubin (GB200 cluster system) production delays, confirms accelerator is in production.

Nvidia CEO Jensen Huang pushed back against reports of manufacturing delays for the company's next-generation Vera Rubin AI accelerator platform, stating that production is already underway. Speaking with reporters at a developer event in Tokyo on Wednesday, Huang dismissed the claims, saying "Vera Rubin is already in production. Giant amounts of production incoming," according to Bloomberg.

His remarks came in response to a report from research firm SemiAnalysis indicating that Nvidia's Vera Rubin AI server rack system, including its Kyber rack platform, faced manufacturing challenges centered on a specialized circuit board used to connect electronic modules. SemiAnalysis is widely regarded for its analysis of Nvidia's supply chain and AI infrastructure developments.

Nvidia had previously stated through a company spokesperson that the Vera Rubin platform remains on track and its roadmap has not been pushed back. The SemiAnalysis report had raised investor concerns that delays could provide competing advantages to rivals such as Advanced Micro Devices, giving the company additional time to advance its MI350 and MI400 AI accelerator offerings.

Huang's comments marked the company's first direct public response to the delay allegations, with the CEO emphasizing that production volumes are ramping as planned. Nvidia shares were down approximately 1% in late-morning trading on Wednesday.