Cerebras launches the CS-4, its first multi-wafer system, though the chip inside is not new

Cerebras introduced the CS-4, its first multi-wafer AI inference system combining three WSE-3 Turbo chips, delivering up to 30 times faster performance than GPUs with improved power efficiency. The system uses existing chip design clocked higher, offering 750 petaflops sparse FP16 compute and supports models over 50 trillion parameters.