Cisco Systems, Inc. (CSCO.US) launches new AI network chip! Targeting the large data center market, directly challenging Broadcom Inc. and NVIDIA Corporation.

Cisco Systems, Inc. (CSCO.US) has released a new type of networking chip designed to accelerate information transfer within large data centers. This product may directly compete with products from Broadcom Inc. (AVGO.US) and NVIDIA Corporation (NVDA.US). Cisco Systems, Inc. stated that its new Silicon One G300, a 102.4 Terabits per second switch chip, can power AI clusters at gigawatt scale to support training, inference, and real-time AI workload, while maximizing the utilization of graphics processors and improving task completion time by 28%. The Cisco Systems, Inc. Silicon One G300 chip will be used to drive the all-new Cisco Systems, Inc. N9000 and Cisco Systems, Inc. 8000 systems, which are designed for hyper-scale cloud providers, emerging cloud service providers, sovereign clouds, service providers, and enterprise customers. The company announced that the Silicon One G300 chip, systems based on this chip, and accompanying optical devices will begin shipping this year. The company specifically highlighted that the new systems offer a 100% liquid cooling design combined with new optical technology to help customers increase energy efficiency by nearly 70%. Additionally, Cisco Systems, Inc. has enhanced its data center network architecture, known as Nexus One, to allow enterprises to operate their AI networks more easily on-premises or in the cloud. "As AI training and inference continue to scale, data movement is key to efficient AI computing; the network itself has become part of the computation. It's not just about faster GPUs, the network must provide scalable bandwidth and reliable, congestion-free data transfer," said Martin Ronda, Executive Vice President of Cisco Systems, Inc.'s General Hardware Group. He added, "The Cisco Systems, Inc. Silicon One G300 driving our new Cisco Systems, Inc. N9000 and 8000 systems, provides high-performance, programmable, and deterministic networking experiences, enabling every customer to fully utilize their compute resources to scale AI securely and reliably in production environments." Last month, NVIDIA Corporation introduced its next-generation AI computing platform, Vera Rubin, which includes several key networking and infrastructure components. In June 2025, Broadcom Inc. began shipping its Tomahawk 6 series switch chips. Meanwhile, Cisco Systems, Inc. also announced a series of features to help enterprises securely adopt AI technology while maintaining the integrity and control of AI interactions. These features include: * AI Bill of Materials: Providing centralized visibility and governance for AI software assets (including model context protocol servers and third-party dependencies) to ensure AI supply chain security. * MCP Directory: Discovering, inventorying, and helping manage risks across public and private platforms for MCP servers and registry centers to strengthen AI governance. * Advanced algorithm red team testing: Expanding the scope of AI security assessments. * Real-time protection for intelligent agents: Ensuring the security of intelligent agents and applications.