TLDR: Cisco has introduced its new 8223 router, powered by the Silicon One P200 chip, designed to significantly boost AI data center networking with a throughput of 51.2 terabits per second. Concurrently, Google has launched Gemini Enterprise, a new platform aimed at centralizing agentic AI access for businesses, replacing its previous Agent Space app. Both developments underscore a growing industry focus on robust AI infrastructure and accessible AI tools for enterprise.
In a significant push to bolster artificial intelligence infrastructure, Cisco has unveiled its cutting-edge 8223 router, featuring the powerful Silicon One P200 chip. This new hardware boasts an impressive throughput of 51.2 terabits per second, marking a major advancement in AI data center networking. The system is specifically engineered for hyperscalers and large data center operators, addressing the demanding requirements of distributed AI workloads for secure, high-capacity, and energy-efficient connections across metropolitan regions. Cisco emphasizes that the 8223 router will enable customers to scale AI clusters more efficiently by supporting multiple optical formats like OSFP and QSFPDDD, crucial for linking geographically dispersed data centers. Rakesh Chopra, a Cisco fellow, stated, ‘The 8223 enables customers to build large-scale disaggregated fabrics for AI infrastructure with unmatched control and performance.’
Cisco’s recent 2025 AI Readiness Index further highlights the challenges and opportunities in AI adoption. The report indicates that a staggering 54 percent of companies believe their current infrastructure cannot scale to meet the rising demands of AI workloads, with only 15 percent reporting their networks as ‘flexible or adaptable’ enough for the new AI era. The index also found that while 83 percent of surveyed organizations intend to develop or deploy AI agents, and 40 percent expect these agents to be working alongside humans within a year, only 31 percent feel prepared to control and secure agentic AI systems. Jeetu Patel, Cisco’s President and Chief Product Officer, commented, ‘This year’s Cisco AI Readiness Index makes one thing clear: AI doesn’t fail, readiness fails.’ He added that ‘The most AI-ready organisations – the Pacesetters from our research – prove it. They’re four times more likely to move pilots into production and 50 percent more likely to realise measurable value.’
Also Read:
- Red Hat AI 3 Unveiled: Empowering Enterprise AI with Scalable Production Inference and Agentic Capabilities
- Salesforce Bolsters Agentforce AI Capabilities Through Enhanced Partnerships with OpenAI and Anthropic
Meanwhile, Google has introduced Gemini Enterprise, a new platform designed to be the ‘front door’ for agentic AI access within the workplace. This paid service is a successor to last year’s Agent Space app and aims to centralize AI capabilities for businesses. Gemini Enterprise allows users to interact with their enterprise data, conduct AI-powered searches, and deploy agents to automate various business tasks. Built upon Google’s advanced Gemini large language models, the system supports media tools such as Imagine VO and Gemini 2.5 Flash, sometimes referred to as Nano Banana. Thomas Kuran, Google Cloud CEO, described Gemini Enterprise as ‘the new front door for AI in the workplace,’ emphasizing its chat-based interface that empowers business users to discover, build, and interact with AI agents across their organization from a single, unified platform.
