While there is still a lot of discussion about defining ‘cloud computing’ there seems to be a general consensus that it’s a burgeoning market, which will, at some unknown point in the future, hit the requisite ‘bust’. This was my sense this past week after attending not just one, but three separate web operations, cloud computing, and infrastructure events where cloud services were front and center:
I may try to review these conferences in a follow on posting, but mostly I wanted to talk about the ‘cloud hype’, it’s ‘boom’ and inevitable bust.
Before Web 1.0 was Network 1.0 I remember pre-Web 1.0, that era right before the Internet exploded. Roughly 1990 to 1996. This was the era in which network ubiquity was in sight, but nowhere near achieved. I worked at InterNex Information Services, which was one of the earliest Internet Service Provider pioneers in the SF Bay Area, along with many other known (Netcom), and unknown (TLG, CRL, GeoNET) providers. For lack of a better term I will call this Network 1.0.
Yes, there were websites, but in ‘93/’94 they weren’t in heavy commercial use. There weren’t any online pet food purchasing services. High traffic websites sat behind T-1 circuits and only the very earliest examples of Web 1.0 sites, like TheSpot.com (an InterNex customer) existed. For the most part ‘the problem’ that we were solving was just getting people connected.
Network 1.0 vs. The Cloud Shift I like to talk about Network 1.0 and compare it to the emergence of cloud services. The main reason is that it looks like that kind of fundamental shift to me. Network 1.0 never ‘crashed’. There wasn’t a bust per se. The network never got smaller. Yes, some players died, much consolidation happened, but the Internet never stopped growing. Not even for an eye-blink. The value proposition was too large and the momentum too huge. It was a ground shift.
Compare this to the business models built on top of Network 1.0. Web 1.0’s bust was a killing ground for bad ideas and bad business models, but it didn’t kill the net. It just kept evolving. This is because the Internet, and infrastructure in general, is more important than what rides on top of it.
The shift to cloud services is this same kind of fundamental change. The businesses and ideas that ride on top of it may win or lose. Some cloud service providers will surely lose and many will consolidate, but cloud computing services are not a fad and can’t die. The economics, the value prop, and the momentum are just as they were with Network 1.0.
The parallels are not exact between the two, but are very very similar. I don’t have a lot of empirical evidence to prove this as we’re currently in ‘land grab’ mode and no one wants to tell anyone else what’s happening, but I’ve been told in confidence from some EC2 alternatives that they are seeing 30% month over month growth rates in uptake. That’s huge. That’s the kind of growth we saw during Network 1.0.
Unkillable Clouds So what does this mean? The bottom line is that cloud computing services are here to stay. There will be winners and losers as always, but this next generation of infrastructure is going to be a key enabler of new and important business models. It also means we are in for a wild ride over the next several years and that business models that predict accurately (or just get lucky) about what the future holds are likely to create some big winners.
Footnotes  To provide some perspective, in 18 months we grew from a 6 person company with a single Point-of-Presence (POP) to an 80+ person company with 80 POPs in California and a nationwide backbone to 3 of the 5 Network Access Points (NAPs) at the time (late ‘95). Roughly 6 new hires, 100+ new circuits, and 4 new POPs every month.