Cisco introduced a new 51.2 Tb/s Silicon One G302 switching chip on October 12 for top-of-rack AI networking. The chip uses 200G SerDes, supports Cisco 1.6T optics, and is designed to integrate with direct-to-chip liquid cooling.
In its OCP 2026 announcement, Cisco said the G302 is built for top-of-rack switches that aggregate server traffic before sending it into the wider data-center fabric. Cisco plans to begin shipping the chip before the end of 2026. The company also said its 102.4T G300 is now sampling and the first G300-powered systems are shipping.
The announcement fits Cisco’s larger push toward faster Ethernet fabrics for AI. BitcoinVersus.Tech recently covered Cisco and NVIDIA’s planned 12.8 Tb/s USFP optics, Cisco’s rack-scale work with Supermicro and NVIDIA, and Broadcom’s 102.4 Tb/s OCP networking push.
Independent coverage from Unite.AI notes that Cisco is positioning the G302 as part of a broader open AI infrastructure stack spanning switching silicon, optics, SONiC and multi-site fabrics. The key point is simple: as GPU racks get denser, the switch at the top of each rack has to move more data while fitting inside tighter power and cooling limits.
BitcoinVersus.Tech Editor’s Note:
We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate using the QR code provided
BitcoinVersus.Tech is not a financial advisor. This media platform reports on financial subjects purely for informational purposes.
Leave a Reply