Looking at the Past to See the Future

In early May 2011, I wrote a blog post in which I dubbed the last ten years the “lost decade” for networking and venture backed networking startups.  The last ten years were not all bad.  Google went public and there were a number of networking IPOs starting with RVBD in 2006 and a handful recently such as BSFT, but none of these networking companies are going to be the next CSCO, JNPR, CIEN, ASND, etc.  My supposition is that none of them are solving a big enough problem and changing the rules by which the incumbents play.  I think the world of APKT, but I am not sure they are going to force some companies out of business.  I think it is more likely they grow for a long time and eventually get bought by a large incumbent who capitulates; realizing they will never catch APKT.  I followed my Lost Decade post with a post on the end of the era for large scale venture capital backed innovation.

On Friday, a colleague sent me a link to this article from Lightreading.com circa October 2000.  The article starts with this premise “Over the next few years, a growing number of carriers are likely to take the plunge into deploying so-called all-optical switches — ones that steer light pulses among different fiber spans without converting them into electrical signals at any point.  The attractions of using all-optical switches are significant. They promise to relieve potential capacity bottlenecks, slash costs, and make it easier for telecom operators to deploy future developments in transmission technology. But reaping these rewards means taking some big risks. In particular, the switching fabric that will form the basis of most all-optical switches is at an early stage of development.”

Here we are in late August 2011, almost ten years later and very few all optical networks have been deployed.  I left the networking industry at the end of 2006 because it was really boring.  The innovative companies, some of which went public in the last few years, were not really that innovative.  Most of these companies are not solving large, difficult problems that create enormous opportunities for wealth creation.  From my perspective, they were just fixing inefficiencies in the network and admittedly some of them did that in an innovative manner – but their solution set lacks scale.  I view the last five years as a period of passing the time until the factors align themselves that allow for really big architectural changes to occur in the network.   I am now convinced we are cusp of the start of that period change.

This is a completely provocative blog post, but I think there are number of forces converging that really call into question what companies are going to be the dominant networking companies for the next 10-20 years.  I think the next 15 year period is the super cycle to play for as number of forces and markets start to interact.  I cannot be 100% confident of the result of the interactions, but I do know that when new companies attempt to (i) do something different, (ii) do something bold and (iii) change the economics of the solution, big changes can happen.  Some will fail, but the winners will force the incumbents of today into the background like DEC, Apollo, Data General, Wang, Silicon Graphics and long trail of discarded tech companies.

I think have outlined my thinking on the evolution of the network in prior posts.  If I was asked to summarize it in a paragraph, I would say we have spent the last ten years laying the foundation for a transformational series of events in the domain of networks.  The big drivers are:

1.      Fiber: More fiber to an ever decreasing unit of users.  Fixing the access portion of the network by driving fiber deeper to the consumer as well as the enterprise is huge change.

2.      Virtualization: Improved CPUs and I/Os capable of supporting 50-100 VMs.

3.      Capacity Upgrades: 1 to 10G to 100G.  The 10G to 100G upgrade in the DC and the service provider network will have a huge impact.

4.      PCs to Tablets and Consumption Habits: In just 18 months we have seen the beginning of the demise of the PC market.  Maybe you think I am wrong, but how are those netbooks selling?  Oh yea…we do not hear much about netbooks anymore.  HPQ is looking to Ten years ago HP announced a $25B acquisition of Compaq.  Now that the Lost Decade is over, HP wants to sell it.  That is a sign that we are exiting the Lost Decade and changes are afoot.

5.      Mobile Devices: Not long ago just getting a mobile phone was an evolutionary step and if you carried a device like a Blackberry you must be really important because you had email in your hand.  Today, who cares?  Everyone has email on their mobile device (I prefer to use my iPad) and most the planes I fly on today have wifi.  The consumption habit induced by my iPad is vastly more than my Nokia feature phone from 1995.

I see optical networks emerging in parts of the network where they have never been before and all optical networks back on the discussion table for the first time in decade.  I think incumbent suppliers are starting to reconsider the architecture of their next generation solutions.  That is the only conclusion I can gather from this Network World article.  Another qualitative data point is I have found a handful of startups (some early, some late stage) looking to solve complex and difficult network problems.  What is interesting about these companies is they have one venture backed competitor at the most, and the majority of them have no venture backed competitors.  That means they have a free run at the incumbent suppliers.  As a veteran of four startups, I like those odds.  I would like to be the small player versus 3-4-5 large incumbents with no rival startup.  If you told me there were 3-5 startups in space then that is a red flag.  I have rambled on too long for a hurricane Sunday.  Just some thoughts and I could be wrong.

/wrk

* It is all about the network stupid, because it is all about compute. *

** Comments are always welcome in the comments section or in private.  Just hover over the Gravatar image and click for email. **