Networking

08:06 AM
Connect Directly
Twitter
LinkedIn
Google+
RSS
E-Mail
50%
50%

Why IBM May Abandon SDN

The news that IBM may sell its software-defined networking division seems to fly in the face of reason. Why exit a burgeoning market? The answer may be more obvious than you think.

IBM often sells off segments of its business before profit margins fall off a cliff. The most well known example was the ThinkPad division sale to Lenovo in 2005. Even though IBM had strong ThinkPad sales, a robust product lineup, and a healthy following in enterprise organizations, business leaders felt the time was right to exit the market.

Today, IBM is making similar moves. The company recently sold its x86 server division to Lenovo, and the latest word around the water cooler is that Big Blue is seeking buyers for its semiconductor business and its SDN business. I can understand the desire to exit semiconductors. That market is fairly well saturated, and profit margins have become razor-thin over the past few years.

Software-defined networking (SDN) is another matter. The technology has been credited with the ability to completely transform enterprise IT. The potential for growth is said to be enormous -- and we are only in the infancy of development. So when IBM seeks to dump its SDN technologies for $1 billion, it should raise some eyebrows. Considering that SDNCentral projects the SDN market to reach $35 billion by 2018, a $1 billion sale is peanuts. 

I can think of three possible reasons why IBM might want to sell its SDN division. First, the company could feel that its product is not on par with competitors. I have serious doubts about this theory, because the technology is still too fluid to be able to start picking favorites. IBM has some amazing engineers, and it's likely their SDN architecture would rival the best offerings from other vendors. 

[Need to learn the basics of SDN? Check out our slideshow, "7 Essentials of Software-Defined Networking."]

A second possibility that many point to is the fact that IBM appears to be shedding all things hardware. Instead, IBM wants to be a software and services-oriented company. This in fact was stated in IBM's press release as the rationale for the sale of its server division to Lenovo, so it makes sense the company would stick to that claim for its SDN division as well.

But to me, the idea that IBM only wants to sell software and services doesn't pass the smell test. The company can't be totally averse to hardware. In fact, it's still one of the best hardware manufacturers in the world. So why would IBM give up on the potential of grabbing a sizable portion of a $35 billion market? Likely because it doesn't believe profitability will be near the levels of hardware markets of the past.

Remember, SDN is all about flexibility. Many architects are drawn to SDN precisely because it is open source and can operate on generic hardware. The concept of mating proprietary hardware and software together is often rejected these days. Customers don't want to get locked into a particular vendor. And when technology becomes a commodity, as is the case with open systems, profit margins plummet.

So while the SDN market may indeed flourish, profit margins will be lower because competitors can only compete based on cost. It's my belief that IBM recognizes this, and the risk/reward equation simply doesn't make sense for the company. I have no doubt that SDN will be the next big evolution in enterprise IT, but look for many of the typical players -- including IBM -- to bow out and seek areas where profitability opportunities are brighter.

Interop Las Vegas, March 31 to April 4, brings together thousands of technology professionals to discover the most current and cutting–edge technology innovations and strategies to drive their organizations' success, including BYOD security, the latest cloud and virtualization technologies, SDN, the Internet of things, and more. Attend educational sessions in eight tracks and visit an Expo Floor more than 350 top vendors. Register with Discount Code MPIWK for $200 off Total Access and Conference Passes. Early Bird Rates end Feb. 21. Find out more about Interop and register now.

Andrew has well over a decade of enterprise networking under his belt through his consulting practice, which specializes in enterprise network architectures and datacenter build-outs and prior experience at organizations such as State Farm Insurance, United Airlines and the ... View Full Bio
Comment  | 
Print  | 
More Insights
Comments
Oldest First  |  Newest First  |  Threaded View
SaneIT
50%
50%
SaneIT,
User Rank: Apprentice
2/18/2014 | 8:48:01 AM
When to sell
I can't say if it's wise for IBM to sell off now but I do think that SDN is going to be a very crowded market in the next few years and it will be difficult to make money.  The open standards and the fact that so many of the big players are making interoperable systems it's going to be hard to lock someone into a single vendor solution and migrating to another vendor should be very easy for your second or third generation project.  Maybe IBM can become the next Xerox PARC, that would be very cool to see happening again.
Drew Conry-Murray
50%
50%
Drew Conry-Murray,
User Rank: Strategist
2/18/2014 | 9:36:24 AM
Dump Switches, Focus on SDN Applications and Services?
This is still speculation since IBM hasn't confirmed its plans. However, I could see IBM getting out of the network hardware market without getting out of the SDN business. The company could focus on SDN applications and services, which would make more sense for IBM. To my knowledge, IBM didn't have all that much market share for data center switches, so it makes sense the company would shed that business along with servers.

At the same time, IBM is heavily involved in OpenDaylight, which makes an open-source controller. IBM could continue to develop applications that can interact with OpenDaylight and other controllers, and also offer integration services for companies trying to roll out SDN projects.
RobPreston
50%
50%
RobPreston,
User Rank: Apprentice
2/18/2014 | 10:36:14 AM
Re: Dump Switches, Focus on SDN Applications and Services?
Has IBM ever shown a long-term commitment to the networking business? It sold its networking hardware unit to Cisco more than a decade ago before getting back in. 
@mbushong
100%
0%
@mbushong,
User Rank: Moderator
2/18/2014 | 11:57:15 AM
Re: When to sell
I don't think SDN is a market so much as an architecture. There will not be any real net-new spend because of SDN. It simply shifts existing network spend to a new class of product. There will, of course, be a change in whether dollars are spent on hardware or software, but I view that more as a change in pricing mix than a material change to networking at large.

That said, you have to wonder who will integrate SDN architectures in existing companies. There will be an opportunity for professional services for systems integrators. That integration will happen at the controller level more than the hardware. I wonder how IBM views that opportunity, particularly if it encroaches on the rest of IT. 

-Mike Bushong (@mbushong)

Plexxi
Andrew Froehlich
50%
50%
Andrew Froehlich,
User Rank: Strategist
2/18/2014 | 1:47:41 PM
Re: Dump Switches, Focus on SDN Applications and Services?
"The company could focus on SDN applications and services, which would make more sense for IBM"

I think you're absolutely correct @Drew. IBM seems to think that software and support is the way to go...
Andrew Froehlich
50%
50%
Andrew Froehlich,
User Rank: Strategist
2/18/2014 | 1:49:34 PM
Re: Dump Switches, Focus on SDN Applications and Services?
@RobPreston -- But SDN is far more than just networking. It's storage, management and monitoring of the entire data center. So while in some ways, IBM isn't involved directly with infrastructure networking, they are involved heavily in other parts of the data center including SAN and infrastructure management tools.
Charlie Babcock
50%
50%
Charlie Babcock,
User Rank: Ninja
2/18/2014 | 8:30:49 PM
IBM savvy about open source
I think Andrew is guessing correctly. IBM has a sense that either Open Daylight or the SDN part of OpenStack or both have staying power, and therefore software-defined networking will start out as a commodity market, with thinning margins from there. IBM is using the vitality of open source code as a predictive power in the marketplace. And IBM, unlike some companies, has always been saavy about open source.
jgherbert
50%
50%
jgherbert,
User Rank: Ninja
2/28/2014 | 10:20:41 PM
Re: When to sell
@mbushong:

I don't think SDN is a market so much as an architecture. There will not be any real net-new spend because of SDN. It simply shifts existing network spend to a new class of product. There will, of course, be a change in whether dollars are spent on hardware or software, but I view that more as a change in pricing mix than a material change to networking at large.

Mike is wise (Hi Mike ;-). Seriously, by definition SDN must be able to keep costs under control or it isn't a truly viable alternative unless it brings to the table something staggeringly fantastic that can offset the additional cost, just like any new technology. Mike also mentions the ingegration of SDN with legacy networking, which is a key point; it's not like the whole world will suddenly become an SDN shop. Greenfield deployments might be that way, but brownfield deployments will likely be little ponds of SDN in a wider legacy network, slowly growing as confidence grows with it. So while the SDN market value is going up, I wonder what's happening to the "legacy networking" market forecasts at the same time if our dollars are going on controllers and software rather than intelligence in the network hardware? 
Slideshows
Cartoon
Audio Interviews
Archived Audio Interviews
Jeremy Schulman, founder of Schprockits, a network automation startup operating in stealth mode, joins us to explore whether networking professionals all need to learn programming in order to remain employed.
White Papers
Register for Network Computing Newsletters
Current Issue
2014 Private Cloud Survey
2014 Private Cloud Survey
Respondents are on a roll: 53% brought their private clouds from concept to production in less than one year, and 60% ­extend their clouds across multiple datacenters. But expertise is scarce, with 51% saying acquiring skilled employees is a roadblock.
Video
Twitter Feed