Cisco Introduces G300 AI Networking Chip, Competing with Broadcom and Nvidia

Cisco Systems (CSCO) introduced the Silicon One G300 switch chip, designed to accelerate data flow in large data centers and compete with offerings from Broadcom and Nvidia in the $600B AI infrastructure spending boom. The G300, manufactured using TSMC’s 3nm process, features "shock absorber" capabilities to reroute data around traffic spikes, improving end-to-end network efficiency. Cisco expects some AI computing tasks to see up to 28% faster performance by automatically re-routing data microseconds in response to bottlenecks. The chip is expected to ship in the second half of 2026.