The computer industry started out with one massive generalist: IBM. However, after facing anti-trust pressures, Big Blue was forced to open up the market.
Specialists like Microsoft, Oracle, EMC, and Cisco entered the scene, with each corporation becoming dominant in its own right. But then EMC told Dell “no” too often, Microsoft went into hardware with the Zune and Xbox, Cisco started to build servers, Oracle bought Sun and suddenly specialists went out of vogue.
Meanwhile, HP invested heavily in software and bought Palm for WebOS, which it now plans to take to other vendors. HP also snapped up 3Com to flesh out its portfolio and adjust to the new model.
Subsequently, both Dell and Intel decided to go their own way with networking, while Cisco is reminded yet again why the path it took towards servers may not have been the wisest one.
High end switches carry a huge margin – much like high end servers back in the days of yore when they were largely proprietary and before x86 penetrated the space. Cisco has traditionally owned the market and it is their success in this lucrative space which largely made the company.
Nevertheless, Intel has reiterated that it is fully capable of doing to high end switches what it did for high-end servers: turn the low and mid-range offerings into something far more affordable. Meaning, higher volumes and lower margins – or the same move that contributed heavily towards putting Sun Microsystems out of business.
Intel’s Fulcrum acquisition is their first big step towards realizing this scenario and it will likely be the end of high margin switches except at the very high end – much like Intel took out all but the highest end proprietary servers. Of course, both HP and Dell could be the big beneficiaries of such a move.
When Cisco went into competition with HP and Dell, there was every likelihood one or the other would respond. As you may recall, HP fired first with 3Com. Since then, HP has pounded Cisco’s margins and market share, with the upcoming 6,500 person layoff perhaps the most salient result. In analyzing this from a macro standpoint, there is likely little benefit for Cisco to be involved with servers, even as it attempts to recover some of the losses incurred from HP aggressively encroaching on their territory.
Well, this week the other shoe dropped when Dell bought Force10 and now intends to go aggressively into the space they once partnered in with Cisco. Force10 boasts a strong, though limited, product line which has been hurt by the small size and reach of the company. Nevertheless, once the deal is approved, Dell should be able to address the Force10 shortcomings on day one.
The end result? There will be another major entry into the business switch market. While HP can be expected to primarily target the enterprise and Dell the mid-market, the combination is likely to place additional pressure on Cisco. Indeed, the two firms will perceive – at least for networking – a common and more lucrative target than each other.
In short, by going into servers, Cisco turned both HP and Dell from at least partial supporters to major competitors. Consequently, the additional pressure placed on Cisco’s volumes and margins should be significant.
Wrapping Up: Fulcrum and Force10 – Oh My!
Intel maintains close relationships with HP and Dell. Both realize Intel is likely to use the Fulcrum technology to bring low cost high performance switches to market. And obviously, you see the bigger problem for Cisco.
To be sure, the market could flip much like it did for x86 servers. Yet, because the space is already prepped for such a move (no one is going to argue Intel’s technology isn’t enterprise ready this decade like they might have in the 1990s), it is more likely to shift rapidly to the new architecture. Remember, switches tend to come as a bundle. Unlike servers, they aren’t expected to run third party applications, just integrate into network management structures – which both Dell and HP know how to write to.
Essentially, rather than the transition slowly occurring over a decade like it did with Sun and x86 servers, this one could happen over just 12 months, with viable products from both HP and Dell hitting the market before this time next year.
As such, Cisco will need to take a hard look at what happened to Sun and immediately work to avoid the same outcome. Few companies can, or have done this, as most seem to ignore historical precedence.We’ll know shortly if Cisco repeats the same fatal mistake.