Amazon.com’s (AMZN.O) cloud services unit said this week that it has not halted any orders of Nvidia’s (NVDA.O) most advanced chip on the market, but has instead decided to purchase the company’s newer state of the art chip for an upcoming supercomputer project between the companies.
Earlier, the Financial Times had reported that Amazon Web Services (AWS), the world’s largest cloud services firm, had “fully transitioned” its previous orders for Nvidia’s Grace Hopper chip to its newer Blackwell graphics processing units (GPUs), that Nvidia announced in March.
An AWS spokesman told Reuters that the transition from Grace Hopper chips to Blackwell chips applies only to Project Ceiba, a supercomputer that AWS and Nvidia are building together. AWS continues to offer other services based on Nvidia’s Hopper chips, its flagship model for training artificial intelligence systems (AI), the spokesman said.
The Project Ceiba transition was announced by the two firms in March when Nvidia unveiled the new Blackwell chips.
“To be clear, AWS did not halt any orders from Nvidia. In our close collaboration with Nvidia, we jointly decided to move Project Ceiba from Hopper to Blackwell GPUs, which offer a leap forward in performance,” an AWS spokesman said in statement.
The Financial Times amended its earlier story to say that Amazon’s chip orders had not yet been placed. The newspaper directed a request for comment to its updated story.
