Cerebras introduces CS-4 wafer-scale AI system with up to 30x faster inference

Cerebras launched the CS-4, its fourth-generation wafer-scale AI system built around three Wafer Scale Engine 3 Turbo processors and a new modular Nexus rack architecture. The company claims up to 30 times faster inference than GPU systems, more than 1,000 tokens per second on models exceeding 10 trillion parameters, and up to 10 times higher throughput per watt versus the prior generation. The system supports disaggregated inference and is designed for rapid deployment at hyperscale, with first shipments beginning this quarter.