Cisco Blogs

Cisco’s Unified Fabric in the Data Center

February 24, 2011 - 5 Comments

Our industry certainly loves its buzzwords.  For a while, everything was about “virtualized this” and “virtualized that” in the data center.  Then there is a current love affair with “cloud”.  It seems the next hot buzzword might turn out to be fabric.

For Cisco, “fabric” in the data center has defined our data center strategy and vision for the last three years.  With the introduction of the Cisco Nexus family in January 2008, we also announced the concept of Unified Fabric as a fundamental building block for the data center.  We offered the simple vision of a single fabric to link all the network, compute and storage resources in a data center as a mechanism to not only reduce TCO but also improve agility and flexibility.  Since then, we have released a steady flow of products and technologies to deliver on the promise of Cisco Unified Fabric by simplifying the infrastructure with convergence, improving its ability to handle virtual and physical scale and increasing the intelligence of the fabric to increase agility and lower operating costs.

While initially hesitant, customers and industry experts are beginning to see the merits of Cisco’s vision, especially in the age of virtualization and cloud.  Meanwhile, other vendors in the marketplace are left to play catch-up.  In a November 4, 2010 independent report titled “Q&A: Networking Landscape, Q4 2010” Forrester Research, Inc. commented that:

“To Cisco’s credit, it saw the data center evolution way before any other networking vendor and started to build a set of products and solutions directed at a converged and virtual world.”- Forrester Research, Inc.

Customer acceptance of our model can be gauged by the remarkable traction we have accomplished with our data center portfolio including with over 10,000 Cisco NX-OS customers with 7M Nexus ports shipped (1M of which are 10GbE) and over 4,000 Cisco UCS customers. When Gartner polled attendees at its recent Gartner Data Center conference about perception of vendor delivery of a fabric-based strategy, respondents listed Cisco first (see figure 2).

In my experience, I think one of the reasons our customers turn to us is because we are experienced at transforming the data center.  Over the decades, we have helped customers converge multi protocol networks on IP, converge IBM SNA onto IP, collapse voice and video onto IP and now, incorporate storage onto the same network.  In each case, we have not just introduced technology, but also provided the expertise to help customers successfully navigate the transition.  The approach has been evolutionary, and I truly believe a huge piece of this is our unwavering commitment to investment protection.  Out entire strategy is build around the ability to incrementally add capability and capacity in a granular as needed.  The Cisco Catalyst 6500 (first shipped in 1999) and Cisco MDS 9500 (first shipped in 2003) are prime examples of this approach.  Today, Cisco continues to reject the one-size fits all approach and the notion that leveraging new innovation means automatically rip-and-replace. In a new report titled “Fabric Computing Poised as a Preferred Infrastructure for Virtualization and Cloud Computing Gartner research Vice Presidents George J. Weiss and Andrew Butler urge customers: “”Do not overhaul or plan on rip-and-replace fabric development, but impose on vendors’ design guidelines that harmonize their solutions with your other data center infrastructure.”

We could not have said it better ourselves.

In an effort to keep conversations fresh, Cisco Blogs closes comments after 60 days. Please visit the Cisco Blogs hub page for the latest content.


  1. Cisco having its presence in every market , every nook and corner of the world, if it keeps its eyes open shall see whats coming ahead of its time. Having money generates lot of marketing collateral around it. One size fits all and all the time is built into Cisco DNA. For e.g IOS in every platform for close to 2 decades, Swiss Army knife of 6500, CRS are great examples.

  2. CISCO is very innovative with the fabrics and they are a great team on the job,nice!

  3. Dear John,

    I would suggest to re-brand IT as “FabrIQ” with the “>” on the “I” as a CISCO (TM) as it delivers “Sophisticated INTELLIGENT Services” and the IQ with > on the I = INTELLIGENCE.

    Kindest regards,

    Jamel E. SEGHAIER
    Premier Multi-Million Linux & Open-Source Entrepreneur MENA
    Started in 1998 as One-manshow with Micro-credit & Big-dream

  4. John, nice post. I also concur with the same Gartner report you mention that states customers should alwyas have 3 vendors qualified for every pat of their infrastructure and avoid single-vendor lock-in. The type of architectures that these proprietary fabrics build.

    I am very glad to see that Cisco does not believe in proprietary fabric architectures. (except maybe for the CRS, right?)

    I also love that Cisco continues to innovate in its own silicon in the data center and not use the merchant silicon other vendors are falling into. It will be a sad day indeed if Cisco sells out the intellectual property of owning its own silicon designs.


  5. From the cloud to fabric, eh? Or will it be The Fabric when it goes mainstream? The concept certainly makes sense, but mainly for large organizations, I would think. Smaller outfits have the valuable advantage to actually do rip/replaces when necessary. Though the evolutionary approach does have a certain appeal…