Powered by the introduction of the brand new Silicon One P200 chip, Cisco’s groundbreaking 8223 routing methods redefine safe, environment friendly AI networking – enabling seamless ‘scale-across’ architectures that join AI clusters throughout a number of knowledge facilities.
-
Cisco’s new AI networking methods redefine what’s doable with unprecedented scalability, energy effectivity, and programmability constructed to instantly tackle the vital challenges of connecting a number of knowledge facilities to securely run AI workloads.
-
The 8223 routing methods ship industry-leading capability and effectivity in a single ASIC router, and at the moment are transport to preliminary hyperscalers for safe, scalable AI infrastructure.
-
Cisco Silicon One is powering the following technology of AI networking with the brand new P200 chip, delivering deep-buffer routing silicon and enabling interconnect bandwidth scale of over 3 Exabits per second.
Cisco unveiled the Cisco 8223, the {industry}’s most optimized routing system for effectively and securely connecting knowledge facilities and powering the following technology of synthetic intelligence (AI) workloads. As AI adoption accelerates, knowledge facilities face hovering demand, rising energy constraints, and evolving safety threats. The Cisco 8223 rises to the problem as the one 51.2 terabits per second (Tbps) Ethernet fastened router constructed for the extraordinary site visitors of AI workloads between knowledge facilities. Cisco at present additionally introduced its newest Silicon One innovation – the P200 chip – which sits on the core of the 8223. Collectively, these improvements empower organizations to shatter bottlenecks and future-proof their infrastructure for the AI period.
“AI compute is outgrowing the capability of even the biggest knowledge heart, driving the necessity for dependable, safe connection of knowledge facilities a whole bunch of miles aside,” stated Martin Lund, EVP, Cisco’s Widespread {Hardware} Group. “With the Cisco 8223, powered by the brand new Cisco Silicon One P200, we’re delivering the large bandwidth, scale and safety wanted for distributed knowledge heart architectures.”
Additionally Learn: AiThority Interview with Tim Morrs, CEO at SpeakUp
In most of the world’s knowledge facilities, AI workloads are stretching energy and area limitations. Hyperscalers can’t scale-up (add extra capability into every particular person system) or scale-out (join a number of methods inside an information heart) any additional. This dynamic places growing demand on knowledge heart interconnects, because the {industry} should “scale-across” by distributing AI workloads throughout a number of knowledge facilities. With out addressing the connection factors between knowledge facilities, organizations might face efficiency challenges, capability bottlenecks and suboptimal processing that wastes time, energy and cash. The Cisco 8223 system offers organizations the flexibleness and programmability needed to construct these networks, with deep buffering to offer the cross-site safety and reliability needed for essential workloads.
Energy Environment friendly. Scalable. Programmable. Safe.
Constructing a future-ready spine AI community at present is crucial – one that may meet the ability consumption challenges whereas remaining scalable, versatile and safe. The Cisco 8223 offers clients the capability to deal with surging workloads, the required flexibility with totally programmability, and the ability effectivity to instantly tackle energy consumption challenges.
The 8223 is uniquely:
- Energy Environment friendly: The 8223 is probably the most energy environment friendly routing system for scale-across networking. It’s a deep-buffer routing resolution optimized for fastened deployments that gives switch-like energy effectivity, instantly addressing the excessive power calls for of AI workloads. As a 3RU system, it’s the most area environment friendly system of its variety. As AI clusters proceed to ‘scale-across,’ energy and area effectivity will solely develop in significance.
- Scalable with out Compromise: The 8223 provides industry-leading bandwidth and the very best density of any fastened routing system within the {industry}. The one fastened routing system that includes 64 ports of 800G, the 8223 provides unmatched routing efficiency able to processing over 20 billion packets per second and scaling as much as 3 Exabits per second. It additionally options 800G coherent optics assist, enabling knowledge heart interconnect and metro purposes reaching as much as 1000km. With the P200’s deep buffering capabilities, the brand new routing methods can take in large site visitors surges from AI coaching, sustaining efficiency and stopping community slowdowns.
- Clever and Adaptable: The 8223 can intelligently adapt to real-time community situations. With the good and programmable P200 silicon, it could assist new, rising community protocols and requirements with out requiring pricey {hardware} upgrades. Networks can stay agile whereas stopping efficiency bottlenecks and accelerating the adoption of recent options as AI site visitors continues to evolve.
- Safe: The 8223 provides safety in any respect ranges – throughout {hardware}, software program and full networks. With options like line-rate encryption utilizing post-quantum resilient algorithms, built-in safety safeguards and steady monitoring instruments, the 8223 can safeguard towards rising threats. Seamless integration into Cisco’s observability platforms offers clients granular insights into community efficiency to assist determine and resolve points rapidly, making certain AI knowledge site visitors is safe and dependable.
Flexibility – An AI ‘Should Have’
The calls for of AI networking aren’t merely rising, however are additionally continually altering and evolving. Organizations demand infrastructure that’s agile sufficient to vary as enterprise necessities do, and are searching for flexibility in deployment fashions to allow them to construct networks for his or her precise wants. The Cisco 8223 will initially be accessible for open-source SONiC deployments, with IOS XR on the horizon.
Along with its availability within the fastened 8223 system, the P200 silicon itself will likely be deployable in modular platforms and disaggregated chassis, providing clients architectural consistency for any dimension community. The Cisco Nexus portfolio will even assist methods working NX-OS primarily based on the P200 within the close to future. Cisco is delivering the agility and operational flexibility required for at present’s AI infrastructure.
Cisco Silicon One: The Trade’s Most Scalable and Programmable Unified Networking Structure
Cisco Silicon One is an entire portfolio of networking units throughout AI, hyperscaler, knowledge heart, enterprise and repair supplier use circumstances. Launched in 2019, Cisco Silicon One is enjoying vital roles in main networks all over the world.
Additionally Learn: Cognitive Product Design: Empowering Non-Technical Customers By means of Pure Language Interplay With AI-Native PLM
[To share your insights with us as part of editorial or sponsored content, please write to psen@itechseries.com]
