"The UALink Consortium wants to create an alternative to Nvidia's interconnect that works with any accelerator and matches the jolly green giant's performance. The group thinks emerging 'neoclouds' that specialize in hosting AI systems will appreciate the chance to build one interconnect capable of handling any GPUs they deploy."
"Version 1.0 of the UALink spec appeared in August 2025. The consortium published version 2.0 today. The big change is the new 200G Data Link and Physical Layers (DL/PL) Specification, which splits the UALink Common Specification by creating one workstream for the group's protocol and the transport layer, and another for I/O tech."
"As explained to The Register by UALink Consortium chair Kurtis Bowman, this means the group can build for the 200G networks of today, the 400G networks that will soon be available, and whatever comes next at the physical layer."
The UALink Consortium, comprising major tech companies, is developing GPU networking standards to compete with Nvidia's NVLink and NVSwitch. New specifications have been released, including version 2.0, which introduces a 200G Data Link and Physical Layers Specification. This aims to support diverse GPU fleets and enhance performance. The consortium envisions a collaborative ecosystem similar to Ethernet, allowing various vendors to create compatible products. The focus is on building an interconnect that can handle the demands of emerging AI systems and future networking technologies.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]