Cisco Blogs

Innovation: It’s in our very core

September 21, 2016 - 21 Comments

There is greenfield innovation. And then there is innovation that shakes you out of your own comfort zone. Without comparing the two, our story is steeped in the second. There was an era when Cisco’s enterprise products not only looked very different from each other but also operated on completely different operating systems. These products were generally built in parallel, from the ground up, and were often tailor-made for specific segments. What this meant was that although the products worked effectively, they weren’t standardized. The lack of standardization was especially apparent when one examined the numerous different operating systems that existed. Innovation is about finding new ways to deal with pain points, and we decided it was time to fix this particular one.

We wanted all our enterprise hardware platforms to run on a unified software stack. Which is why we created a modernized IOS XE software stack for all of our enterprise routing, switching and wireless products, internally code named Polaris. This groundbreaking common software stack allows us to offer a completely consistent user experience across our entire portfolio while at the same time using a common platform to continue our industry-leading innovation. Releasing features has become vastly more efficient as they can be developed in a single release train across all enterprise platforms. And, as we continue to build next-generations platforms and capabilities we can do so with hitherto unmatched engineering efficiency.

Single Pane of Glass

So what does this mean for the future?  We are on a path to making enterprise-grade software as intuitive and adaptive as consumer technology. For the first time in the history of Cisco, we have a single software stack across the enterprise routing and switching product lines. Polaris not only brings together a single software stack across enterprise, it also lays a modern infrastructure foundation. A foundation that will help define enterprise networks of the future by enabling automation and programmability for network management and a slew of advanced functionalities like serviceability, device and network analytics, smart licensing, and more. For instance, streaming telemetry, built natively and deep into the software stack, will allow network data from the devices to be consumed and processed by controller layers to create information and actionable insights for new forms of innovation in network monitoring and control. We will also bring together wired and wireless functionality together for seamless operation and consistent policies across wired and mobile devices. Just as for smartphones and other consumer devices, we can make available an app store-like experience for enterprise third party apps to run on the network devices for fog computing.

A common software stack for the enterprise is the way of the future. It allows for increased efficiency, ease-of-use, consistency in operation as well as new innovation and much more.  At the end of the day, however, delivering the best possible product and experience to our customers is what drove this re-invention of enterprise networking.  Onward and forward!

Look forward to continuing the discussion @aoswal1234.


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. From IOS to MCP to IOS-XE in only eight years truly is a remarkable journey. Once capable of running in a few devices and core routers, IOS-XE is bulked up like a Sumo and ready to party in a vast range of Enterprise-grade devices.

    With a greatly increased product footprint and a feature set of infrastructure and APIs that customers have never dreamed of having, this Sumo is Cisco’s champion in the marathon race to total market domination. All this and it is still managed in with the same familiar way customers know and love!

    Forward-thinking architecture and modern tooling are sure to enable remarkable developer productivity, and we will all enjoy the fruits of these continuing efforts in accelerated release velocity.

    Congratulations on Polaris!

  2. Also lol @ “Cisco Social Rewards.” Exactly the “me too” delusions of grandeur and relevance that a company with a functional innovation engine would not do.

  3. IOS XE was a very fulfilling project and aligns well with Chuck’s recent announcement about how spin-ins may be back on the menu (since XE was out of the BCN spin-in). It would be nice, however, to Cisco extend this kind of innovation to other parts of the business.

    UCS has gotten not much love lately; what’s interesting about it Security is abandoned. Cloud.. ? Not happening.

    The problem as I see it is that Cisco has these enormous cash cows that even without serious investment deliver huge revenues year after year after year. The risk to a business like Cisco is that an inflection point comes along and turns it all on its head.

    Unlike others I appreciate that Cisco is not yet in that position; there are storm clouds on the distant horizon (things like cloud are likely to change the business for Cisco but not nearly as soon as people think) however.

    The real question people should be asking is why so many Cisco acquisitions go nowhere. Meraki was a great model for next generation CPE but isn’t really happening. Cisco has spent a ton of money on “cloud” stuff that doesn’t seem to be going anywhere (cliqr, metacloud, jasper, piston, parstream), ditto security (embrane, tailf, …)OpenDNS is a great acquisition but how are you taking it forward? Cariden was best in the business but where is the synergy? etc.

    In other words, it sure seems like the innovative core ran out of juice in the 200xs and now the reactor is just going cold.

    Innovation at Cisco has a long, long history of BUY(Aironet, Crescendo, S1, Cerent, Growth, etc.!)-or-SPININ(nuova, bcn, insieme, …). Those are great. But lately the buy list looks pretty grim; it’s like a bunch of stuff that’s on clearance and which are “safe” (boring) buys.

  4. That’s it, you said it :…”increased efficiency, ease-of-use, consistency in operation as well as new innovation.”…delivering the best possible product and experience to our customers is what drove this re-invention of enterprise networking. Onward and forward!. I think that Cisco is always delivering a great “innovation” and “consistency”in every product. Thanks for sharing your post and insights.

  5. Excellent blog post on the cross platform standardisation with unified software, Anand. Such innovations would certainly augur well for your company to build a solid foundation for even bigger and better innovations to come. Improved resource optimisation, swift trouble-shooting and easy amalgamation of future features, ……. this would definitely be a game changer ! All the best Anand, to you and Cisco

  6. Sorry guys, but this is not innovation. Other vendors like Juniper has JUNOS OS runnig on their entire porfolio, not only in Enterprise equipments. Definitely is abig step for Cisco!

  7. Thank you Mr. Oswal for clearly articulating key points of this software/standardization/OS transition toward a more intelligent, efficient and effective future. Harmony is a good thing.

  8. a) Polaris originally supposed to ship about 2 years ago.

    b) Lovely message, if you don’t have any NX-OS in your estate.

  9. I feel there are still some limitation. More in terms of using standard opensource software used by data-center/devops like docker, logstash, puppet chef ansible etc. I mean why not something similar to cumulus Linux ?

  10. First time in Cisco’s history, Polaris had enabled single software stack running across routing, switching and wireless platforms. This has not only simplified customers operations but have greatly improved engineer’s daily life working on these products. Polaris foundation and DNA vision has great potential to change future networks are going to be built. Great to be part of this journey!

  11. I am from tooling.
    Similar to what is done in Polaris. I would love to see in tooling. All Software development tools needs to be made part of one stack.
    also under one group ( One leader )

    The software development tools in Cisco are very complex.
    I see a need from Leaders to push towards forming a common software stack for Cisco’s Development tools as well.

    Having a common stack for all software development tools will
    – Make the development and maintenance of tools more efficient.
    – Make the tools well integrated.
    – Will make the overall software development more efficient.

    By software development tools, I am referring to
    – Repository management tool such as ACME, and many other related one.
    – Build tools ( Make file )
    – Smoke test/CI tools S2S etc.
    – and many others.
    – Build tools,

  12. Great post. Proud to be a part of such a innovative and forward looking team with the great leaders.

  13. Nicely articulated Anand. It was great and ambitious initiative.Our feature velocity, productivity is greatly improved. The culture of more than one team for same work doesn’t exist now. Thanks to unification made by Polaris . From low-end TSN to high end ASR1K Polaris has build tremendous foundation. You can invent when foundation is solid , we have rock solid foundation now. Innovation will reach to new hight now. Great work and thanks for your leadership Anand .

  14. ‘non-greenfield’ innovation. Real tricky but immense impact!!!

  15. A thought provoking article on linking innovation to solve real problems. How do we reap benefits of the common software stack – Polaris – to enable machine learning, software monetization, unified interface to the customer would be key for future success.

  16. One software stack simplifies customer deployment, awesome work to achieve this. Not too often such a large exercise is done in the industry. Good work team.

  17. The level of efficiency this brings to Cisco is HUGE!!!
    One software stack to bring innovation into.
    One software stack to support, address issues.
    One software stack to develop and test on.
    One software stack to address vulnerabilities on, to bring secure technologies into.
    One software stack for all NMS (APIC-EM, WebUI, PnP, etc) to interact with.
    All of this, across all platforms converged into that stack, the amount of effort saved, the focused effort this enables, this will keep building and becoming more evident over the years to come. The converged focus of all teams into one software stack, the quality will go through the roof!
    And this is only on the efficiency! So much more beyond efficiency that this convergence will bring!

  18. I completely agree that enterprise software does not have to be geek’s software. The next generation operators will likely prefer the software that does not impose/expose complexity, works like a charm- this even consumer software struggles with today despite being much simpler. they would prefer network to auto detect failures and update in network as much policy driven and intelligent as possible. Bottom line users won’t want to use their intelligence in figuring out how to use software, rather what they can get out of software by using it.