A 5% improvement in power supply efficiency will help save energy; so will incremental improvements in air conditioning and power handling and storage systems and computer processors. Every item in the data center should be as efficient as you can get. That’s a common sense idea and a more common factor in buying decisions.
That’s not enough.
I noted last week that the new Hitachi array boasts a 20% energy efficiency gain but that it isn’t enough.
And it isn’t.
But even the substantial efficiencies gained by EMC are not enough. And the efficiencies gained by every competitor and every other supplier of data center infrastructure are not enough either.
Every bit will help, but no bit is enough. No single component will save or break the energy budget. And the problem is just too big for power supply improvements or small improvements in a small storage array to have enough impact. Suppose you take an integrated approach that looks at servers, storage, networks and software together? Suppose you also include the facility itself, with air conditioning and power handling?
All good. Every opportunity should be taken to improve these. Still not enough.
There isn’t a silver bullet here. And if there were, it probably wouldn't be made from technology. The more customer conversations that I have on the topic, the more I have come to believe that changes in organizational dynamics and individual behavior are way more important than advances in technology. Technology helps, but technology by itself doesn’t change behavior. At VMworld a few weeks ago I observed two things that helped to illustrate this for me.
John Chambers, CEO of Cisco, gave a keynote at VMworld. He held everybody’s attention for more than an hour in a room of ten thousand conference attendees. His talk was fascinating for his very specific look at the future of technology and its impact on how we will live and work.
Behavior change was part of it.
Among his topics, he contrasted a recent series of web-enabled Cisco business meetings against a three week round-the-world business trip he had taken the year before. (Cisco makes a device called Telepresence that is used for realistic virtual meetings.) Even with a few technical glitches, his assessment was that he successfully accomplished in a few days what it had taken three weeks of exhausting travel to accomplish, less successfully, the previous year. He stayed off the airplanes, reduced his personal impact on the environment, and set an example to change the behavior of Cisco employees. He also got in a little product promo for their Telepresence device.
What struck me was what happened next.
Following the Chambers presentation was an excellent panel discussion about data center energy efficiency. By the time the panel completed, about half the room had emptied.
Why?
I can only speculate that those who didn’t stay had no interest and didn’t see how the topic had an impact on them and their jobs. Perhaps they figured the boss was taking care of it. Perhaps they didn’t realize how much impact each individual can have. Perhaps they need more examples that show them how.
The EPA report released earlier this year pointed out that as much as half of data center energy consumption could be eliminated using existing technology and best practices. So it seems that it isn’t the technology that is deficient but rather the way it gets used. And that’s a matter of behavior. It's also a matter of good examples and information and education.
Will new technologies help? Yes. Especially because IT people are very busy, working very hard to get the job done and looking for every means to do it successfully. If we in the industry - EMC and its partners and its competitors too - will do a better job at illustrating how to do that, then those busy people will be more energy efficient and more successful too.
The great thing about energy efficiency as a topic is that you soon come to the conclusion that it really is business efficiency that you get. Energy efficiency is just a very important side effect.
Comments