Nvidia has to go back to the drawing board for Blackwell after chips overheated in server racks. Customers are starting to get worried.

Blackwell is a concern for Nvidia. After an initial design flaw, Nvidia is now struggling with overheating. The chips can overheat when placed in high-performance server racks with up to 72 processors, knowledgeable sources say. The risk of overheating drastically reduces the performance of the chips.

Concerned customers

Nvidia is now working with its partners to adapt the design of the server racks so that the Blackwell chips feel comfortable there. The company assures that this is part of the “normal development process”. But The Information is hearing mixed voices from customers who have some concerns about whether Nvidia will be able to deliver high-performance Blackwell chips on time.

Previously, Nvidia had to recall the first Blackwell prototypes due to a design flaw that led to low yields. Jensen Huang himself admitted that the problem was “one hundred percent Nvidia’s fault.” At the end of October, the hoped-for news came that Blackwell had received the green light for large-scale production and the first deliveries were expected in January. It is unclear whether this schedule will be affected.

Waiting impatiently

The big players in the AI ​​and cloud industry are eagerly waiting for Blackwell. Companies like Google, Microsoft and Meta bought up all of Blackwell’s shares before the first chip even rolled off the assembly line. The tech industry announces even greater investments in AI after every quarter.

Nvidia benefits from this like no other company due to its dominant market position. Huang can now even call himself CEO of the most valuable company in the world.