What Cisco’s SDN Spin-In Move Tells Us

Many of you have followed a series of posts I’ve written on rumblings that Cisco’s renowned engineering troika  of Mario Mazzola, Luca Cafiero, and Prem Jain would be reuniting to launch another venture.

Rumors last summer suggested that they might be incubating a networking company, perhaps in conjunction with a Valley venture capitalist. Subsequent rumors indicated that the Cisco engineering trio was building a switch as part of a startup company, maybe even as part of another Cisco spin-in company.

Spinning Back

During the last two weeks, rumors intensified and suggested that the threesome was building a data-center switch attuned to the requirements of cloud computing. It also became clear that this would indeed be another Cisco spin-in company. Now we learn, from a report in the New York Times, that the switch in question will feature software-defined networking (SDN), and that the principals behind the spin-in venture, called Insieme — it means “collection” or “assembly” in Italian — are involved in business negotiations with Cisco.

We don’t know much other than that, though. When asked by the New York Times about Insiemi, Cisco CEO John Chambers invoked a cone of silence, saying “we do not discuss our plans or internal investments.”

Well, hey, somebody’s been discussing this particular plan, if not the specific investment terms pertaining to it, because this has not been a particularly well-kept secret. Information has leaked out about it, some of it perhaps intentionally, for some time.

The negotiations relating to Insiemi will be about remuneration, deliverables, and timelines. Cisco will tie compensation to the realization of specific objectives. Now that it has come this far, getting reported in the New York Times, I doubt that it will go back into mothballs. It’s doubtless moving ahead.

Messages Imparted

So, what does that tell us?

Well, it tells us a few things. First, it indicates that Cisco felt it again needed the services of its spin-in wrecking crew, the team that came to Cisco initially in its first-ever acquisition, of Crescendo Communications back in 1993. That brought Cisco the Catalyst line of switches, which was no small prize, along with a talented roster of personnel that played a significant role in the company’s growth into an industry giant. After coming to Cisco, Crescendo’s engineering stars — they would be in Cisco’s Hall of Fame, if such a thing existed — then were involved in Cisco’s highest-profile spin-in efforts: Andiamo for storage networking, and Nuova, which developed data-center technology that found its way into Nexus switches and UCS blade servers.

That Cisco felt it needed the spin-in touch, especially involving this particular group of engineers, also tells us implicitly that Cisco didn’t feel the job could be done by the teams it already has working inside the company, including David Yen’s Server Access and Virtualization Technology Group (SAVTG). That’s interesting in and of itself, because Yen came over from Juniper Networks to effectively take the reins from Mazzola, Cafiero, and Jain, who then transitioned to “support Cisco in an advisory capacity.”  In that advisory capacity, which they assumed last spring, the trio reported to John Chambers directly, not to Yen’s bosses, senior vice presidents Padmasree Warrior and Pankaj Patel.

Potentials Risks As Well As Rewards

And now they’re back in the spin-in saddle, and you can make of that what you will. Rest assured, however, that much will be made of it on the Cisco campus . . . which brings us to the third thing that this move tells us.

These spin-in moves are not universally popular within Cisco Systems. While Cisco had entirely valid business and technology reasons for instituting its spin-in model, the practice has generated much internal discord and friction. Cisco employees not chosen to participate in the spin-in ventures have been known to become alienated and invidious. (I suppose “pissed off” might sum it up, but we usually aim for a higher order of decorum and eloquence around here.)

That was one of the reasons that I wondered, back in the spring of 2010, whether Cisco might have retired its spin-in move. While some external observers contend that Cisco overpays for its spin-in ventures, Cisco insiders who don’t get to travel on the spin-in express aren’t pleased about left behind on the station platform. In 2008, former Cisco executive Jayshree Ullal, who now serves as CEO of Arista Networks (more on which later), made the following comment to Forbes about the malignant consequences of spin-in ventures:

“Spin-ins are a creative model to accelerate innovation and bring in engineers you couldn’t normally recruit–and financial gains go to entrepreneurs, not venture capitalists,” says Jayshree Ullal, a 15-year Cisco veteran who built the 7000 then left last May as the Nuova people came back in. “But it’s a nightmare when the guy in the next cubicle is a multimillionaire and you aren’t, because you weren’t chosen.” She left Cisco for personal reasons, she says, adding that she had to deal with a lot of unhappy employees over the spin-in structure.

Cisco Takes SDN Threat Seriously

So, there will he happy employees and unhappy ones at Cisco, those who get tapped for the not-so-secret spin-in society and those who get left behind to maintain the workaday business. How troublesome that becomes, and whether it results a new stream of defections, remains to be seen.

One of those previous defections involved the aforementioned Jayshree Ullal, now CEO of Arista Networks. I intimated above that Arista figured into this story, and it does, as do Nicira Networks and the new breed of SDN purveyors.

If Cisco is betting big on a Mazzola-Cafiero-Jain spin-in venture related to SDN — and past performance tells us that these ventures are never small wagers — it tells that the Cisco takes very seriously the threat posed in the data center by Arista, which has staked its own SDN ground, and by SDN startups such as Nicira.

Cisco’s conception of SDN, as fashioned by its spin-in wrecking crew, might diverge in interesting ways from those others have put forward. Watch how terms are defined, and who does the defining, as the battle for hearts, minds, and wallets intensifies.

26 responses to “What Cisco’s SDN Spin-In Move Tells Us

  1. Where does Embrane.com fit in then? That company is also led by former CSCO folks and appear to building a L4-7 SDN???

  2. Does not speak highly of Cisco’s confidence in their Nexus product line, generally these spin-ins have been done to acquire new technology not replace existing…

    • Mark B what are you talking about nexus confidence?? We just finished doing a bakeoff in my company between nexus, qfabrIc and Arista. And let tell you. Qfabric was joke, then the final 2 companies were Cisco and Arista. We decided to go with Cisco. Nexus was more stable and solid code than the Arista 7508 and the 7050 could not half of the stuff that Arista account team promised. Yes we ran into small issues with nexus but the account team was all over that something that Arista could not do. Plus Arista was pitching their new low latency to find out from another source in the last minute that alta has been delayed. Bottom line Cisco still Cisco.

      • Do you work at Cisco? If not then how are you speaking to their confidence level in the Nexus platform?

        The fact Nexus is reaching max fabric capacity with the latest fabric upgrades and the F2 line card has no ability to to 40/100G – requires a separate VDC and external cables to connect it to the M-series modules to leverage all of the great features seems like a perfect recipe for a new high end switch focused at 100G not an SDN play.

        I am glad you picked the switch that was best suited for your environment I did not once bring up vendors or products but glad you did as it is always good hear a success story. I am completely vendor neutral in any post I make not trying to bash anyone just asking questions that many do not and always looking for more from vendors – not taking what they have and settling with good enough.

        Each vendor as strengths and weakness as demonstrated by this spin-in move Cisco does not believe they can innovate internally so they are having their ‘A team’ go out and build a new platform.

  3. Nexus is very Enterprise focused where SDN is all about Cloud so I can see them coexisting just fine.

    • Simon – You should tell that to all of the cloud providers Cisco is pushing Nexus in, or are you saying nexus can’t support those customers and Cisco is selling an inferior solution?

      I am trying to figure out with all the smart people in Cisco why they have to pay off others to ‘innovate’ in their core routing/switching market…

  4. Pingback: Cisco incubating software-defined network startup? | GLC SX MM

  5. Pingback: Cisco Incubating Software-defined Network Startup? | Geeklin

  6. Pingback: Cisco Incubating Software-defined Network Startup? | BiZZBo@rd | Find People Who Share Your Business.

  7. Pingback: Cisco Incubating Software-defined Network Startup? | DIGITAL VELOCITY

  8. Pingback: Cisco Incubating Software-Defined Network Startup?

  9. If it is true that Insiemi is an SDN play, the delicious irony is that the people in charge are a bunch of hardware guys who tolerate software as a necessary evil that allows them to peddle their hardware.

    May God have mercy on the souls of the poor software devs who jump on board.

    • If I am not mistaken its Insieme (together in Italian) while they are hardware types I would never underestimate ability to deliver a solution. With SDN controllers in general my bigger question are people trading CapEx savings on hardware for CapEx spend on software (in some cases it would be OpEx). Regardless is there true savings or is it shifting who gets the money. In the end something needs to be done to simplify data center management/operations. Great example is the Nexus 5000 / Nexus 2000 solution while technically it is a step back to the days of the Cat6000 and MSM but it is popular (aside from having the Cisco sales machine behind it) because it does provide some management benefits. Its time to look at how we do things and figure out to implement an ‘easy button’ right now SDN and their controllers seem to be the hot new thing that will save us – while I am very optimistic only time will tell…

  10. Pingback: Cisco Incubating Software-defined Network Startup? | The Software Site

  11. Pingback: Cisco possibly incubating software-defined network startup | Datacentre Management . org

  12. Pingback: New Spin-In Called Insiemi: A Framing Exercise « SIWDT

  13. Mark B’s comments seem valid. Asking if he is Brad does is not an answer to the points he brought up. At the end of the data Total Cost of Ownership wins. In this case we are moving towards the idea that single management style deployment works better then the existing designs. That is what SDN is really about. How do I as company X scale my network for “big data” and manage all the complexity that comes at scale? That is the question to be answered. The companies that answer that problem will be the ones that win in the market place.

  14. Mark B i would think is Mark Berly of Arista. As always he makes good points, but he should be clear about who he works for.

  15. Hi Brad,

    In my personal opinion, John Chambers “spin-in strategy” is an “end-run” around federal securities laws to evade Cisco’s Board of Directors corporate compensation guidelines and its up to Cisco employees to report this to the SEC.

    I mean, what kind of Board of Directors allows scandalously exorbitant corporate compensation shenanigans on such an epic scale?

    According to the Cisco Board of Director’s Compensation Committee, to carry out its purpose, the Committee has the following responsibility and duty:

    “Review annually and approve the Company’s compensation strategy to ensure that it promotes shareholder interests and supports the Company’s strategic and tactical objectives, and that it provides appropriate rewards and incentives for management and employees of the Company, including review of compensation-related risk management.”

    Make no mistake about it, Mario Mazzola, Prem Jain and Luca Cafiero are still most definitely Cisco employees hiding under the “disguise” of a “spin-in” called Insieme.

    So why do I believe Mario Mazzola, Prem Jain and Luca Cafiero are still most certainly Cisco employees?

    Well, starting a company involves financial risk, and these guys have absolutely no financial risk at stake.

    Cisco CEO John Chambers is using his so-called “spin-in” strategy (for the 3rd time no less) to hide from Cisco shareholders as well as tens of thousands of much, much lower paid Cisco engineers, the true and scandalously exorbitant compensation that he’s actually paying to these 3 particular engineers.

    Cisco employees can report their internal knowledge about these guys and their scandalously exorbitant corporate compensation shenanigans to the U.S. Securities and Exchange Commission (SEC) Office of the Whistleblower:

    “The Whistleblower Program was created by Congress to provide monetary incentives for individuals to come forward and report possible violations of the federal securities laws to the SEC. Under the program eligible whistleblowers (defined below) are entitled to an award of between 10% and 30% of the monetary sanctions collected in actions brought by the SEC and related actions brought by other regulatory and law enforcement authorities.

    “The Program also prohibits retaliation by employers against employees who provide us with information about possible securities violations.”

    Note: You are not required to be a Cisco employee to receive a monetary award from the SEC.

    Sincerely,

    Brad Reese

  16. Pingback: Cisco Networkers Live CLUS or whatever we call it now SDN Predictions | | Network StaticNetwork Static

  17. Pingback: IETF Draft - BGP-signaled end-system IP/VPNs

  18. Cisco’s spin-in is legalized corruption. It is as simple as that.

Leave a comment