Apple has Joined A consortium building next-generation technology to link chips together in AI data centers.
A consortium, called the Ultra Accelerator Link Consortium, is developing the standard UALlinkwhich combines the AI accelerator chips found in a growing number of server farms. As of Tuesday, Apple is a board member of a consortium with Alibaba and semiconductor company Synopsis.
In a statement, Becky Loop, Apple’s director of platform architecture, said UALink “shows great promise” in addressing connectivity challenges and creating new opportunities to expand AI capabilities and demands.
“Apple has a long history of collaborating on initiatives and innovations that advance our industry,” added Loop, “and we are excited to join the UALink board of directors.”
UALink aims to connect chips from GPUs to custom-designed solutions to speed up the training, fine-tuning and running of AI models. Based on open standards including AMD’s Infinity FabricThe first UALink products are expected to launch within the next two years.
Intel, AMD, Google, AWS, Microsoft, and Meta are among the members of the UALink consortium. Nvidia, by far the largest manufacturer of AI accelerators, is not. That’s probably because Nvidia offers its own proprietary interconnect tech, NVLink, to link chips in a data center cluster.
Apple’s involvement in ULink comes as the company ramps up its investment in infrastructure to support Apple IntelligenceIts suite of AI product features. According to the Wall Street Journal, Apple is said to developing A new server chip to improve the efficiency of its AI data centers.
Some of Apple Intelligence’s capabilities have met with mixed reviews. (My colleague Sarah Perez is called them “boring and practical.”) Last week, Apple said It will update one of those capabilities, AI-treated news alerts, after it reported seeing inaccurate headlines, including tennis star Rafael Nadal coming out as gay.
https://techcrunch.com/wp-content/uploads/2019/06/GettyImages-1143959321.jpg?resize=1200,791
