AMD and Cerebras announce disaggregated inference collaboration
Cerebras Systems Inc. →AMD and Cerebras announced a technical partnership combining AMD Helios rack-scale systems with Cerebras Wafer-Scale Engine technology in one inference workflow. The planned architecture assigns prompt processing and large-context work to AMD hardware, while Cerebras handles decoding and token generation. The companies presented this separation as a way to address different workload requirements for latency, throughput, scale and cost. Intended applications include coding, real-time copilots, autonomous agents and other interactive systems. Cerebras planned to deploy AMD Helios in its data centers, with initial access to the joint solution expected through Cerebras Cloud during the second half of 2026. The release describes a collaboration and expected availability, not confirmation that deployment had completed.