A group of data center tech leaders have formed the Ultra Accelerator Link Promoter Group to create a new way to scale up AI systems in data centers.
Advanced Micro Devices, Broadcom, Cisco, Google, Hewlett Packard Enterprise (HPE), Intel, Meta, and Microsoft today announced they have aligned to develop a new industry standard dedicated to advancing high-speed and low latency communication for scale-up AI systems linking in data centers.
Called the Ultra Accelerator Link (UALink), this initial group will define and establish an open industrystandard that will enable AI accelerators to communicate more effectively. By creating an interconnectbased upon open standards, UALink will enable system OEMs, IT professionals and system integrators tocreate a pathway for easier integration, greater flexibility and scalability of their AI-connected data centers.
“The work being done by the companies in UALink to create an open, high performance and scalableaccelerator fabric is critical for the future of AI,” said Forrest Norrod, general manager of the Data Center Solutions Group at AMD, in a statement. “Together, we bring extensive experience in creating large scale AI and high-performance computing solutions that are based on open-standards, efficiency and robust ecosystem support. AMD is committed to contributing our expertise, technologies and capabilities to the group as well as other open industry efforts to advance all aspects of AI technology and solidify an open AI ecosystem.”
The Promoter Group companies bring extensive experience creating large-scale AI and HPC solutions based on open standards, efficiency and robust ecosystem support. Notably, AI chip leader Nvidia is not part of the group.
“Broadcom is proud to be one of the founding members of the UALink Consortium, building upon our long-term commitment to increase large-scale AI technology implementation into data centers. It is critical to support an open ecosystem collaboration to enable scale-up networks with a variety of high-speed and low-latency solutions,” said Jas Tremblay, VP of the Data Center Solutions Group at Broadcom.
Scaling for AI workloads
As the demand for AI compute grows, it is critical to have a robust, low-latency and efficient scale-upnetwork that can easily add computing resources to a single instance. The group said Creating an open, industry standard specification for scale-up capabilities will help to establish an open and high-performance environment for AI workloads, providing the highest performance possible.
The group said this is where UALink and an industry specification becomes critical to standardize the interface for AI and Machine Learning, HPC, and Cloud applications for the next generation of AI data centers and implementations. The group will develop a specification to define a high-speed, low-latency interconnect for scale-up communications between accelerators and switches in AI computing pods.
The 1.0 specification will enable the connection of up to 1,024 accelerators within an AI computing podand allow for direct loads and stores between the memory attached to accelerators, such as GPUs, in thepod.
The UALink Promoter Group will immediately start forming the UALink Consortium, expected to be official in Q3 of 2024. The 1.0 specification is expected to be available in Q3 of 2024 and made available to companies that join the Ultra Accelerator Link I(UALink) Consortium.
About Ultra Accelerator Link
Ultra Accelerator Link (UALink) is a high-speed accelerator interconnect technology that advances next generation AI/ML cluster performance. AMD, Broadcom, Cisco, Google, HPE, Intel, Meta, and Microsoftare forming an open industry standard body to develop technical specifications that facilitate breakthrough performance for emerging usage models while supporting an open ecosystem for datacenter accelerators.
“Ultra-high performance interconnects are becoming increasingly important as AI workloads continue togrow in size and scope. Together, we are committed to developing the UALink which will be a scalable and open solution available to help overcome some of the challenges with building AI supercomputers,” said Martin Lund, EVP of the Common Hardware Group at Cisco, in a statement.
The post Data center leaders create Ultra Accelerator Link group for AI connectivity appeared first on Venture Beat.
