Oracle to Offer Cerebras AI Inference in 2027

Oracle will offer Cerebras Systems’ AI inference technology through Oracle Cloud Infrastructure starting in 2027, the companies announced October 27, adding another computing option to Oracle’s cloud portfolio. The planned service will let Oracle Cloud Infrastructure customers use Cerebras systems for inference, the stage when trained AI models respond to requests. The companies said the offering is intended for AI labs, enterprises and developers, including those running applications with multiple steps. Cerebras designs processors using its WaferScale Engine, which places computing resources and memory on a single wafer-scale chip. The company says its architecture can speed model responses. Oracle already offers Dedicated AI Clusters, and the Cerebras technology is being added as another option rather than replacing existing services. Oracle did not disclose how much Cerebras hardware it would use. The companies also did not provide service capacity, pricing or deployment scale. The announcement did not identify customer commitments for the planned offering. Mahesh Thiagarajan, president of Oracle Cloud Infrastructure, described the addition as another option in Oracle’s...