Can Cisco’s new AI data centre router tackle the industry’s biggest infrastructure bottleneck?
Cisco has entered an more and more aggressive race to dominate AI data centre interconnect know-how, turning into the newest main participant to unveil purpose-built routing {hardware} for connecting distributed AI workloads throughout a number of amenities.
The networking big unveiled its 8223 routing system on October 8, introducing what it claims is the industry’s first 51.2 terabit per second fastened router particularly designed to hyperlink data centres operating AI workloads.
At its core sits the new Silicon One P200 chip, representing Cisco’s reply to a problem that’s more and more constraining the AI business: what occurs while you run out of room to develop.
A 3-way battle for scale-across supremacy?
For context, Cisco isn’t alone in recognising this chance. Broadcom fired the first salvo in mid-August with its “Jericho 4” StrataDNX swap/router chips, which started sampling and likewise supplied 51.2 Tb/sec of combination bandwidth backed by HBM reminiscence for deep packet buffering to handle congestion.
Two weeks after Broadcom’s announcement, Nvidia unveiled its Spectrum-XGS scale-across network—a notably cheeky identify on condition that Broadcom’s “Trident” and “Tomahawk” swap ASICs belong to the StrataXGS household.
Nvidia secured CoreWeave as its anchor buyer however offered restricted technical particulars about the Spectrum-XGS ASICs. Now Cisco is rolling out its personal elements for the scale-across networking market, organising a three-way competitors amongst networking heavyweights.
The drawback: AI is just too massive for one constructing
To perceive why a number of distributors are dashing into this area, contemplate the scale of recent AI infrastructure. Training giant language fashions or operating complicated AI techniques requires hundreds of high-powered processors working in live performance, producing monumental quantities of warmth and consuming large quantities of electrical energy.
Data centres are hitting exhausting limits—not simply on accessible area, however on how a lot energy they will provide and funky.
“AI compute is outgrowing the capability of even the largest data centre, driving the want for dependable, safe connection of data centres tons of of miles aside,” stated Martin Lund, Executive Vice President of Cisco’s Common Hardware Group.
The business has historically addressed capability challenges via two approaches: scaling up (including extra functionality to particular person techniques) or scaling out (connecting extra techniques inside the identical facility).
But each methods are reaching their limits. Data centres are operating out of bodily area, energy grids can’t provide sufficient electrical energy, and cooling techniques can’t dissipate the warmth quick sufficient.
This forces a 3rd method: “scale-across,” distributing AI workloads throughout a number of data centres that is likely to be in several cities and even completely different states. However, this creates a new drawback—the connections between these amenities turn out to be essential bottlenecks.
Why conventional routers fall quick
AI workloads behave in another way from typical data centre site visitors. Training runs generate large, bursty site visitors patterns—intervals of intense data motion adopted by relative quiet. If the community connecting data centres can’t take in these surges, all the pieces slows down, losing costly computing sources and, critically, money and time.
Traditional routing gear wasn’t designed for this. Most routers prioritise both uncooked pace or refined site visitors administration, however battle to ship each concurrently whereas sustaining affordable energy consumption. For AI data centre interconnect purposes, organisations want all three: pace, clever buffering, and effectivity.
Cisco’s reply: The 8223 system
Cisco’s 8223 system represents a departure from general-purpose routing gear. Housed in a compact three-rack-unit chassis, it delivers 64 ports of 800-gigabit connectivity—at present the highest density accessible in a hard and fast routing system. More importantly, it could possibly course of over 20 billion packets per second and scale as much as three Exabytes per second of interconnect bandwidth.
The system’s distinguishing characteristic is deep buffering functionality, enabled by the P200 chip. Think of buffers as non permanent holding areas for data—like a reservoir that catches water throughout heavy rain. When AI coaching generates site visitors surges, the 8223’s buffers take in the spike, stopping community congestion that may in any other case decelerate costly GPU clusters sitting idle ready for data.
Power effectivity is one other essential benefit. As a 3RU system, the 8223 achieves what Cisco describes as “switch-like energy effectivity” whereas sustaining routing capabilities—essential when data centres are already straining energy budgets.
The system additionally helps 800G coherent optics, enabling connections spanning as much as 1,000 kilometres between amenities—important for geographic distribution of AI infrastructure.
Industry adoption and real-world purposes
Major hyperscalers are already deploying the know-how. Microsoft, an early Silicon One adopter, has discovered the structure priceless throughout a number of use instances.
Dave Maltz, technical fellow and company vice chairman of Azure Networking at Microsoft, famous that “the widespread ASIC structure has made it simpler for us to broaden from our preliminary use instances to a number of roles in DC, WAN, and AI/ML environments.”
Alibaba Cloud plans to make use of the P200 as a basis for increasing its eCore structure. Dennis Cai, vice chairman and head of community Infrastructure at Alibaba Cloud, acknowledged the chip “will allow us to increase into the Core community, changing conventional chassis-based routers with a cluster of P200-powered units.”
Lumen can also be exploring how the know-how matches into its community infrastructure plans. Dave Ward, chief know-how officer and product officer at Lumen, stated the firm is “exploring how the new Cisco 8223 know-how might match into our plans to reinforce community efficiency and roll out superior companies to our clients.”
Programmability: Future-proofing the funding
One often-overlooked facet of AI data centre interconnect infrastructure is adaptability. AI networking necessities are evolving quickly, with new protocols and requirements rising recurrently.
Traditional {hardware} usually requires substitute or costly upgrades to help new capabilities. The P200’s programmability addresses this problem.
Organisations can replace the silicon to help rising protocols with out changing {hardware}—vital when particular person routing techniques characterize vital capital investments and AI networking requirements stay in flux.
Security concerns
Connecting data centres tons of of miles aside introduces safety challenges. The 8223 contains line-rate encryption utilizing post-quantum resilient algorithms, addressing issues about future threats from quantum computing. Integration with Cisco’s observability platforms supplies detailed community monitoring to establish and resolve points rapidly.
Can Cisco compete?
With Broadcom and Nvidia already staking their claims in the scale-across networking market, Cisco faces established competitors. However, the firm brings benefits: a long-standing presence in enterprise and repair supplier networks, the mature Silicon One portfolio launched in 2019, and relationships with main hyperscalers already utilizing its know-how.
The 8223 ships initially with open-source SONiC help, with IOS XR deliberate for future availability. The P200 will probably be accessible throughout a number of platform sorts, together with modular techniques and the Nexus portfolio.
This flexibility in deployment choices may show decisive as organisations search to keep away from vendor lock-in whereas constructing out distributed AI infrastructure.
Whether Cisco’s method turns into the business customary for AI data centre interconnect stays to be seen, however the elementary drawback all three distributors are addressing—effectively connecting distributed AI infrastructure—will solely develop extra urgent as AI techniques proceed scaling past single-facility limits.
The actual winner might finally be decided not by technical specs alone, however by which vendor can ship the most full ecosystem of software program, help, and integration capabilities round their silicon.
See additionally:
Want to study extra about AI and large data from business leaders? Check out AI & Big Data Expo happening in Amsterdam, California, and London. The complete occasion is a part of TechEx and is co-located with different main know-how occasions together with the Cyber Security Expo, click on here for extra info.
AI News is powered by TechForge Media. Explore different upcoming enterprise know-how occasions and webinars here.
The submit Can Cisco’s new AI data centre router tackle the industry’s biggest infrastructure bottleneck? appeared first on AI News.
