Monthly Archives: September 2009

Moore’s Law and the Cloud Inflection in IT Staffing

I was in a meeting last week with Gartner’s Ben Pring and he made an interesting observation that cloud computing at the end is just a result of Moore’s law.  The concept is fairly simple and charts a path of increasingly distributed computing from mainframes, to minicomputers, to workstations and PCs (which resulted in client/server), then on to the Internet, mobile computing, and finally to cloud computing.  But cloud computing is not an increase in distribution of computing — it’s actually the reverse.  Sure, there are more devices than ever.  But since internet application topologies have replaced client/server, the leveraging of computing horsepower has migrated back to the data center.

The explosion in distributed computing brought on by ever faster processors (coupled by lower prices on CPUs, memory and storage) allowed for the client/server revolution to push workloads onto the client and off of the server.  Today, much of the compute power of edge devices (PCs, laptops and smart phones) is not used for computing, but for presentation.  Raw workload processing is happening on the server to an increasing degree.

Much like a supernova that collapses into a black hole, workload distribution is becoming ever more concentrated in the data center.  But in this case the data center is also being concentrated. The rise of the hosting economy that started with the internet era continues to accelerate.  But the level of concentration of data center resources into a small number of super galactic providers is accelerating even faster.  Amazon and Google are gaining share while much of the traditional hosting and outsourcing community is losing out.  Some may not see it yet because their sales are still increasing, but a short look at some of the analysis by Guy Rosen on Amazon vs. Rackspace, GoGrid and Joyent shows that even some of the larger players out there are losing share to the galactic few.

This concentration is affecting the corporate data center too, of course.  More than half of Amazon’s instances are going to large corporate users according to CTO Werner Vogels.  It’s still very early days, and most large companies continue to invest in data center infrastructure, but you don’t have to be a genius to see that at some point the cloud concentration effect will be felt by even the largest of IT organizations.  No matter how you measure it, cloud computing is going to have a material impact on the number of servers and related components that the average enterprise is responsible for managing.  With the possible exception of storage, the effect will be to shrink the corporate data center.

So what does this have to do with IT staffing?  The galactic-scale cloud providers have invested heavily in a level of automation that most enterprise IT organizations can only dream of.  The result is that the number of servers per systems administrator at Amazon or Google is likely many times (if not orders of magnitude) larger than your typical IT shop.  As more and more of the data center moves to the cloud and consolidates in the galactic core, the need for systems administrators and related IT staff functions will be diminished.

Until the cloud, Moore’s law resulted in a steady increase in demand for skilled systems and network administrators.  At some point, the economies of scale and concentrating effects of cloud computing – particularly in the area of IT operations – will be visible as a measurable decline in the demand for these skills.

This reminds me of a t-shirt I got from 3tera at Interop in May, the back of which reads “Real men don’t manage servers (anymore).”

Disclaimer & New Job

For those of you who didn’t catch it on Twitter, I have recently taken a position at Unisys as the Director, Cloud Computing Portfolio.  Great company, people and opportunity to participate in the cloud evolution taking over this large and storied IT vendor.

The opinions expressed on this blog and in related Twitter, Facebook or other feeds are mine alone and do not reflect the opinions of my employer – Unisys – or anyone else for that matter including but not limited to current, former or future clients, dogs and cats in my neighborhood, or the toad that keeps finding its way onto my porch (ftw?).  Okay?  Okay.

Market Parallels – Cloud and Open Source?

Any new technology market has its own lifecycle and rhythm.  From mainframes, through smartphones, there’s the early years, the rapid growth, some slowing down and inevitably a decline.  Some technologies never go away completely (e.g. mainframes), while others never really get a foothold (insert your own example here).

Open source was a software movement that began as an idea and now dominates how many new software offerings are marketed and sold. Open source is not a technology, but a business and legal framework within which technology is propogated.  Still, the biggest companies in software are largely closed source – Oracle, SAP, etc.  Nearly all specialty vertical apps (e.g. trading systems) are closed source.  Whereas most new development technologies including databases and tools are open source.  Given that open source is more a legal construct which bleeds into sales and marketing, it’s highly likely that there will be both open and closed source models co-existing for any foreseeable future.  Further, open source shrinks the size of the industry from a revenue perspective by default  (though parodoxically, software spending is up this year even in this economy).

What about cloud computing?  Will there still be the need for the cloud modifier in the future?  In the past most infrastructure was sold directly to the users under a capex spending model – this includes servers, databases, operating systems, etc.  Of total infrastructure spending in 20 years, how much will still be for on-premise capex, and how much for cloud opex?  Will ecconomies of scale drive infrastructure in the cloud to a point where the infrastructure market will shrink in both real and nominal terms?

Will the purveyors of servers, networking and core infrastructure software sell 90% of their wares to cloud companies?  Will what we currently call  “cloud computing” be  just plain “computing” in the future?  Time will tell, and it will be a long time before the cloud distinction becomes superflous, but it will be interesting to watch.

Tagged ,
%d bloggers like this: