Anytime you can get a bunch of companies with very technically savvy people with strong opinions working together on a problem or set of problems, that’s when you really know there’s a real problem. Many problems with Ethernet networks at scale prompted the formation last summer of the Ultra Ethernet Consortiumwhich now has over a hundred participating organizations working to deliver its first specification.

UEC’s founding members all have skin in the networking game, and the HPC and AI systems space where Ethernet needs to be supported and expanded to offset the limitations of current Ethernet and InfiniBand fabrics that have traditionally been used for capacity-class systems in the last decade.

Founders include AMD, Arista Networks, Broadcom, Cisco Systems, Atos spinout Eviden, Hewlett Packard Enterprise, Intel, Meta Platforms and Microsoft. Nvidia wasn’t an original member, but recently joined to throw its weight behind some ideas that have been thrown around about how to create Ethernet networks that have multiple terabits per port and can scale to over 1 million endpoints . And in most cases, the endpoint in question is a vector and tensor accelerator that is used to do computations for AI training.

To understand what UEC is all about and how the work is progressing, I had a chat with J Metz, who is the chairman of the consortium’s steering committee. Metz held engineering internships at Apple, QLogic, Cisco Systems, and Rockport Networks before joining AMD as technical director of systems design. He has also served on the boards of the Fiber Channel Industry Association, the NVM-Express Consortium, and the Storage Networking Industry Association, and knows his way around standards bodies that include input from a diverse set of players and experts .

In the first part of this series, I talked about UEC’s mission, breaking down barriers in the OSI model, and the people who maintain the technologies in the different layers of that model and the process of reaching consensus on how to evolve. Ethernet and when different stages are set. We’ve talked before about the things UEC isn’t trying to do, at least not initially, because creating a more extensible Ethernet without breaking backwards compatibility is a tough enough task as it is.

We hope you enjoy the first part of this conversation, and all you have to do is click the video link above and you can hear what Metz had to say.