Media reports that Nvidia’s new Blackwell AI chips have encountered problems with accompanying servers that overheat.
According to information released by journalists last Sunday, November 17, the mentioned problem became a cause for concern for the customers of the specified company. In this case, the concern of clients is based on worries that they won’t have enough time to get new data centers up and running.
Media reports that the Blackwell graphics processing units overheat when connected together in server racks designed to hold up to 72 chips. The journalists published the relevant information with reference to insiders who were aware of the problem.
It is worth noting that Nvidia has asked its suppliers several times to change the design of the racks. In this case, the goal was precisely to eliminate the overheating problem. The relevant information was published by the media with reference to customers and suppliers.
The company spokesperson said in a comment to reporters that Nvidia is working with leading cloud service providers as an integral part of its engineering team and process. It was also separately noted that the engineering iterations are normal and expected.
Nvidia presented Blackwell chips in March. Initially, the company said that these microcircuits would ship in the second quarter. The firm then faced delays that could potentially have an impact on customers such as Meta Platforms, Google, and Microsoft.
Nvidia’s Blackwell chip takes two squares of silicon the size of the company’s previous offering.
It is worth noting that Nvidia is one of the main beneficiaries of the so-called artificial intelligence boom. This year, the company’s market capitalization has crossed the historical mark of $ 3 trillion. The chips developed by Nvidia are necessary for the training and subsequent operation of artificial intelligence systems.
