Overheating accompanying servers have been an issue for Nvidia's new Blackwell AI chips, which have already experienced delays.

According to the information on Sunday, some clients are concerned that they won't have enough time to set up new data centers since Nvidia's new Blackwell AI processors, which have previously had delays, have issues with accompanying servers that overheat.

According to the article, which cited people with knowledge of the problem, the Blackwell graphics processing units overheat when joined together in the specially constructed server racks the business has created.

According to Nvidia workers who have been working on the problem, as well as customers and suppliers who are aware of it, the AI chipmaker has repeatedly requested that its suppliers alter the racks' design in order to address overheating issues, the article stated without naming the providers.

A request for comment outside of regular business hours was not immediately answered by Nvidia.

Prior to experiencing delays that would have an impact on clients like Microsoft, Meta Platforms, and Alphabet's Google; Nvidia had stated that its Blackwell processors would ship in the second quarter when they first revealed them in March.

The Blackwell chip from Nvidia combines two silicon squares the size of the company's last product into a single unit that is 30 times faster at activities like answering chatbot queries.