The European Data Center Summit 2011 was held yesterday at SihlCity CinCenter in Zurich. Google Senior VP Urs Hoelzle kicked off the event talking about why data center efficiency was important both economically and socially. He went on to point out that the oft quoted number that US data centers represent is 2% of total energy consumption is usually mis-understood. The actual data point is that 2% of the US energy budget is spent on IT of which the vast majority is client side systems. This is unsurprising but a super important clarification. The full breakdown of this data:
· 2% of US power
o Datacenters: 14%
o Telecom: 37%
o Client Device: 50%
The net is that 14% of 2% or 0.28% of the US power budget is consumed in datacenters. This is a far smaller but still a very relevant number. In fact, that is the primary motivator behind the conference: how to get the best practices from industry leaders in datacenter efficiency available more broadly .
To help understand why this is important,
· Of the 0.28% energy consumption by datacenters:
o Small: 41%
o Medium: 31%
o Large: 28%
This later set of statistics predictably shows that the very largest data centers consume 28% of the data center energy budget while small and medium centers consume 72%. High scale datacenter operates have large staffs of experts focused on increasing energy efficiency but small and medium sized centers can’t afford this overhead at their scale. Urs’s point and the motivation behind the conference is we need to get industry best practices available to all data center operations.
The driving goal behind the conference is that extremely efficient datacenter operations are possible using only broadly understood techniques. No magic is required. It is true that the very large operators will continue to enjoy even better efficiency but existing industry best practices can easily get even small operators with limited budgets to within a few points of the same efficiency levels.
Using Power Usage Effectiveness as the measure while the industry leaders are at 1.1 to 1.2 where 1.2 means that every watt delivered to the servers requires 1.2 watts to be deliverd from the utility. Effectively it is a measure of the overhead or efficiency of the datacenter infrastructure. Unfortunately the average remains in the 1.8 to 2.0 range and the worst facilities can be as poor as 3.0.
Summarizing: Datacenters consume 0.28% of the annual US energy budget. 72% of these centers are small and medium sized centers that tend towards the lower efficiency levels.
The Datacenter Efficiency conference focused on making cost effective techniques more broadly understood showing how a PUE of 1.5 is available to all without large teams of experts or huge expense. This is good for the environment and less expensive to operate.
–jrh
b: http://blog.mvdirona.com / http://perspectives.mvdirona.com
Thanks for your comment Rocky. As you can imagine, I’m skeptical that you can a new build and operate it for less than the cost of using a cloud service. If you send me your data (jrh@mvdirona.com), one of us will learn something. I’m interested but almost guarantee, AWS can help.
–jrh
I keep looking at the economics of cloud computing for my operations, and keep finding the cloud to be much more expensive. I have no desire to run even a very small data center; but unless I can show my boss that the cloud is cheaper, we’ll keep running a small data center.
I’ve been pressured to reduce the operating costs (increase the power efficiency) of our new, very small data center. Turns out it’s cheaper to build a more efficient data center, because you use less of everything, if you design properly. Someday, I’ll write about what I’ve learned (a lot from this blog), because no one has written a "Dummie’s Guide to Building Small Green Data Centers".
On another angle – certain operations can’t move to the cloud due to bandwidth limits, latency, licensing, and other difficult issues. My ideal is a really small data center for those applications that require local servers, with a very fast connection to a robust and reliable cloud.
So we will always need efficient, low-cost designs for small data centers.
I’ll ask Urs but, intuitively, it makes sense and is roughly what one would expect. Can you point me to the IDC studies?
–jrh
James, interesting stats. Would you have the source referenced for this study/survey? Has different numbers than I’ve seen in previous IDC reports.
Thanks!
Exactly Shawn, that’s the point. Because their are vastly more small centers, they much more power in aggregate than large centers. But, if we don’t get the small centers more efficient, then we won’t move the needle on overall industry efficiency.
Broadly there are two solutions possible if we want to substantially improve overall industry efficiency: 1) make small centers by adopting best practices, and 2) move small centers to the cloud. The former was the goal of the conference. The later has obvious gains. Just use a large operator to host your computing workload. I’m a big believer in the advantages of cloud computing at all scales and expect it is going to be a big part of the answer for operators of all sizes but, scaling down, is particularly difficult.
–jrh
It occurs to me that there are vastly more small and medium data centers than large. I wonder if the energy consumption numbers are artificially skewed smaller for the large centers due to there being fewer of them.
The rough rule of thumb I use are small is under 2MW, medium is 3MW to 20MW to perhaps 30MW. Large runs 35MW and above. Most people start "large" a bit lower — in the 15 to 20MW range. I suspect Urs is using similar cut offs but I haven’t asked him.
–jrh
Was interested in the criteria by which datacenters were categorized as large/medium/small? -Thanks, Anne Holler
100% agree. PUE is less than perfect and, specifically, it says nothing about server efficiency nor s/w efficiency. Its a measure only really intended to be used to measure data center infrastructure (power distribution and cooling) and its far from perfect even at that: //perspectives.mvdirona.com/2010/05/25/PUEIsStillBrokenAndIStillUseIt.aspx.
–jrh
The strength of PUE is that while it’s a good metric of datacentre efficiency, it doesn’t measure server efficiency, how much useful work is done per Joule. If your PUE is 1.2 but your server is so power hungry that even the fans have fans, you are worse off than in a place where the PUE is 3.0 but you are running on laptop parts.
Of course, in an ideal world, you’d have both. Then you can worry about the network power budget, and that between your sites…