Ampere and Qualcomm collaborate on AI chip

Translate from : Ampere og Qualcomm samarbejder om AI-chip
A new AI chip solution from Ampere Computing and Qualcomm is designed to reduce power bills and increase energy efficiency in data centers.

Ampere Computing announced Thursday that it is collaborating with Qualcomm on a new solution aimed at reducing power bills for operating AI chips. Ampere, founded by Renee James, a former president of Intel, uses technology from Arm Holdings to make central processing chips used by Oracle, Google's Alphabet and other companies.

The startup has focused on making chips that are more energy efficient than industry leaders Intel and Advanced Micro Devices. Qualcomm, which dominates the mobile phone chip market, has been working to break into the data center AI chip market since 2019 with an energy-efficient offering of its own. Ampere and Qualcomm announced Thursday that they have integrated their chips into a single data center server.

"Consider this as ... the first of the things we're working on," said Jeff Wittich, Ampere's Chief Product Officer. "Obviously, we can do things that are much bigger than building-level solutions as we go into the future, given that we're both tackling a similar type of problem." The joint Ampere-Qualcomm offering will not directly compete against AI chip leader Nvidia, whose chips are used to train AI systems with huge amounts of data.

Instead, the Ampere-Qualcomm servers are intended to efficiently run these models after they are trained. Both Ampere and Qualcomm also compete indirectly with Nvidia because AI chips are often sold in systems that pair more than one type of chip. The two working together can block potential rivals from gaining traction with customers, said Jim McGregor, founder of Tirias Research.

Our Partners