09 September 2011

Is the Bandwidth Glut Over?

We all remember the internet revolution of the late 90's, but perhaps even bigger was the telecom boom.  Billions of dollars went into internet infrastructure companies that increased network capacity with hopes of trying to cash in on a Netscape-type exit.  Then, the bubble burst.  Fiber lay dormant in the ground.  Investment halted.  Demand didn't materialize fast enough as pundits called it a "glut."   But now, in the midst of internet 2.0, have we past that point and entered into an era of undercapacity?  Many of the projected growth levers of the 90's are finally starting to hit mass adoption in ways never imagined 15 years ago while investment in internet networks has slowed.  With the diverging demand and supply characteristics in play, is the bandwidth glut officially over? Is there a telecom boom 2.0 on the horizon?

The Telecom Act of 1996 helped drive the influx of competition and investment into the market shortly after it was passed. The timing could not have been more perfect as the internet hit the masses around the same time.  Deregulation, lofty projections, and loads of investor cash all created a perfect storm for a boom in capacity.  VCs threw money at companies involved in fiber deployment, electronics and software, and telecom-based real estate all in the haste of meeting forecasted demand.  Unfortunately for those investors, most of them failed. Most of them shut their doors, had their assets sold for pennies on the dollar, or got swallowed up by consolidators like Level 3.

The days of build it and they will come are over.  There are very few investments in pure infrastructure companies right now.  Incumbents such as Verizon and Sprint used to sell capacity through wholesale channels are now scaling back and using it for their own networks.  Once it becomes a cost center versus a revenue source, there is always a risk of underinvestment (remember the IPhone AT&T network fiasco?).  As everyone knows, the economic backdrop in which the current internet 2.0 is now flourishing is cautious at best.  Capital Expenditure budgets are being slashed and scrutinized much more than before.  All of these things are leading to a somewhat supply crunch or at least a lag in deployment.  This is at a time when demand is booming. And will continue to do so for years to come.

During the first wave, people were merely accessing web addresses and text files which required very small usage of bandwidth (excluding Napster and porn).  Now that is a fraction of what we are doing today.  Pictures, streaming video, video conferencing, peer to peer, Facebook, ftp sites like Dropbox are all contributing to a huge increase in bandwidth per user. Cloud computing, a concept that centered on demand forecasts in the late 90's, is finally here in a big way.  But this is just the start.

Nobody back then imagined the worldwide explosion of smartphones and tablets.  All of these users are mapping, chatting, downloading apps, and accessing multimedia via the net on their mobile devices. The problem is the internet networks from the 90's were not built for this demand.  There is a mass rush right now to scale wireless networks to keep pace with demand.  Back then traditional TV and telephone had separate distribution systems.  They have all converged onto IP-based platforms (or moving that way quickly).  Radio is next.  That's huge.  Other mediums and industries that we have not even contemplated are sure to follow further increasing demand for bandwidth.  Just look at the skyrocketing prices of data center stocks like Equinix or Digital Realty or the fact that Amazon's server business is approaching $1B in revenue.

Despite the fact that supply might not be keeping up with demand through traditional means such as fiber-based networks, innovation is on the rise.  Huge strides in wireless networks are being made, new mediums such  power and electricity are becoming more plausible, and component costs continue to drop allowing companies to light those late 90's networks cost effectively.  So while the bubble valuations are back in the web-based world, we haven't seen it in the infrastructure realm.  Since history tends to repeat itself, its only a matter of time before we see Telecom Bubble 2.0.