Is Software Defined Networking (SDN) A Sucker’s Bet?

It’s literally time to step right up and place your bets.  While Software Defined Networking (SDN) is one of the hottest trends in enterprise networking today, it is easy to place the wrong bet.  Looking at the previous large strategic bets that businesses have had to make – the Internet, virtualization, or cloud computing – it was a sucker’s bet.  The answer was always “yes” and the only variables were how much and when.  It was like looking at a blizzard bearing down on the Midwest – you know it is going to hit, but when and how many inches were up for negotiation.  While we have written an extensive analysis here, I want to provide you a snapshot of my thoughts here.

With all of the talk about SDN in the market don’t assume that the answer is a slam-dunk as well.  Vendors are asking enterprise IT to bet on theirvision and follow their path, but the “substitutability” of the products is anything but clear, and picking the wrong vendor can have catastrophic impact on a company’s  business, not only in lost capital and lost productivity, but also in a major opportunity cost.  At a time when the business cycles are starting to fire up again and the economy is looking more positive for the next 12 months than the previous, can you afford to make the wrong call on something this strategic?  Or, to put a finer point on it, can enterprise IT afford to make the call too early?

Following the trajectory of network servers, from mainframe to standard x86, we saw large commoditization in the market.  This was good for end customers and good for vendors. But when the market for commodity hardware started to flatten out, vendors started integrating those standard products into vertical solutions.  It is more profitable to sell a complete stack, and if you can save the customer time in the setup, configuration and integration with a turn-key solution, all the better.

Network switching, typically based on a three or four layer hierarchy, has yet to undergo the same metamorphosis, and more like sublimation is going to go straight from a solid to a gas, skipping that awkward liquid phase in the middle. This is the phase where all of the cost is wrung out and the industry gravitates around a set of commodity building blocks.  The next big thing in network switching will be software defined networks, which promises to bring the agility and flexibility of cloud computing to that rigid and wire-laden world of network switching.

Sounds great, right?  The catch is that SDN, while driven off of a standard called “OpenFlow,” does not represent a commoditization of the market. You will see prices on expensive network switches come down as intelligence is moved from these leaf switches to SDN controllers (which flatten the hierarchy and drive efficiency), but adding the software that makes SDN run will potentially offset those savings.

Different vendors have different strategies, but unlike technology strategies, these may more closely resemble incumbent rather than disruptor strategies. How will each of these vendors stack up against each other?  How do SDN vendors like Cisco, Hewlett-Packard HPQ -0.56%Alcatel-Lucent ALU -0.37%, Juniper, Dell DELL +0.07% compare?  That’s a longer story you can find here, but I’ve placed a quickie comparison below.

Do you agree, disagree?  Let me know.

John Fruehe contributed to this article.