Amazon Web Services (AWS) has clarified that it has not halted any orders for Nvidia’s top-of-the-line Grace Hopper chip. Instead, the cloud services giant has opted to purchase Nvidia’s newest next-generation chip for an upcoming supercomputer project in collaboration with the chipmaker.
Previously, it was reported by the Financial Times that AWS had completely shifted its previous orders for Nvidia’s Grace Hopper chip to the company’s latest graphics processing units (GPUs) Blackwell, which were announced in March. However, an AWS spokesperson told Reuters that this transition from Grace Hopper chips to Blackwell chips is specifically for Project Ceiba, a joint supercomputer project between AWS and Nvidia. The spokesperson clarified that AWS will continue to offer other services based on Nvidia’s Hopper chips, which are known for their performance in training artificial intelligence systems.
The decision to transition Project Ceiba from Hopper chips to Blackwell GPUs was made in mutual agreement between AWS and Nvidia, with the aim of taking advantage of the improved performance offered by the new chips. In response to the initial report by the Financial Times, AWS clarified that no orders were held back from Nvidia, and the transition was a strategic move for the supercomputer project.
Following this clarification from AWS, the Financial Times updated its article to reflect that Amazon’s chip orders had not actually been placed as previously stated. The publication requested comments on the updated story from the involved parties.
The collaboration between AWS and Nvidia on Project Ceiba was announced back in March during the launch of Nvidia’s Blackwell chips. These chips are expected to offer significant advancements in performance, making them an attractive choice for the supercomputer project. AWS and Nvidia remain committed to working together on innovative projects that leverage cutting-edge technology for various applications.
In conclusion, despite earlier reports suggesting a shift away from Nvidia’s Grace Hopper chips, AWS has clarified that the transition to Blackwell GPUs is specific to the Project Ceiba supercomputer project. The collaboration between AWS and Nvidia continues to drive advancements in AI technology and high-performance computing, with a focus on utilizing the most advanced hardware available in the market. This strategic decision reflects the ongoing efforts of both companies to push the boundaries of innovation in cloud computing and artificial intelligence.
Article Source
https://www.aol.com/amazons-cloud-unit-pauses-orders-134219910.html