Green IT -- Are We Missing The Point?

Over time, we're getting more passionate bloggers here at EMC.  And we enjoy debating different topics back and forth among ourselves.  I think we were all argumentative as children -- must be in the DNA.

Dick Sullivan and I were having an interesting discussion that many of us (vendors, IT and business people alike) perhaps missing a really big point around the whole "green IT" thing.

It's simple -- yes, it's important to have "green IT", but what we're really after is efficiency.

And -- more often than not -- the efficiency gains we're after with green IT lead to even bigger efficiency discussions that many times have nothing to do with energy or carbon footprint.

Have You Read Dick's Blog?

Dick came to EMC and IT by way of being an English teacher, which means he writes far better than I could ever hope to.  His blog "Energy Matters" is getting picked up more and more often, which is great, since I think he's doing a great job and a great service in this area.

Dick's been able to evolve a simple, focused discussion (e.g. "green IT") into a systematic approach to sustainability that goes far beyond just IT concerns.  As a result, he's gotten very popular on the speaking circuit as of late.

Dick and I agree on a simple premise -- in the quest for energy efficiency and reduced environmental footprint, we're often led to expose wasteful business practices that might have nothing to do with what we originally set out to find.

High energy use and/or excessive environmental footprint is turning out to be a signature for all sorts of things that ought to be done better. 

The Cisco TelePresence Example

We have Cisco's TelePresence installed between our east coast briefing center and our west coast briefing center. 

For those of you who haven't seen Cisco's offering, this isn't your average video conference room.

Imagine a conference room, cut in half, and three mongo hi-def monitors on the long wall.  The cameras and sound tracking is set up to trick the brain into believing you're really there.  To initiate a meeting, I just press a big, green button -- that's it.

The brain has a funny property -- give it enough "real world" cues, and it wants to fill in the rest.  The fidelity of the experience is so good that -- for many people -- it tips over into being just like a real meeting.

Unlike other technologies I've used, you relax, discuss, pick up on body language and facial cues, and so on. 

Just like a real meeting.

The business case to justify this was based on reduced travel between our two locations.  I don't know whether it's ROI positive or not, but it's staying, and we're probably gonna get a bunch more of these units in the future.

So, one argument that was presented during the justification for this stuff was that the energy differential for using a 2-hour TelePresence session rather than flying to meet in person was a big deal.

On one hand, I'm sure that's important, and I'm glad that people are thinking in these terms.  But really, I think that's missing the big picture.

What's really inefficient is a group of people slogging across the coast, taking two or more days out of their lives, enjoying the pleasures of domestic air travel, just to have a 2-hour face-to-face meeting.

The energy saving / footprint reduction is simply a thin layer of frosting on a pretty beefy cake.  To my point, the energy savings leads you to a wasteful practice, the resolution of which leads to benefits that may far outweigh the proposed energy / footprint reductions.

The Virtualized Data Center Example

One of the big reasons given for moving to VMware is energy savings.  And, yes, collapsing lightly used servers by a factor of 4:1, 8:1 or whatever is a pretty compelling energy reduction, especially at scale.

But, as I've written before, virtualizing your compute environment creates all sorts of resulting benefits that have nothing to do with energy savings or footprint reduction per se.

Provisioning can be done in minutes, rather than days or weeks.  Moving to new platforms (to gain the advantages of new technology) can be done far faster than before.  Disaster recovery and business continuity is far faster and far easier with a VMware-style approach.  And, despite some contrarians, I will argue vehemently that VMware-style approaches to server and desktop are inherently and architecturally more secure than physical alternatives.

I would speculate that the total economic impact of the above benefits (and probably a few I forgot!) might far outweigh any energy / footprint savings.

Once again, a very thin layer of attractive frosting on a very rich cake.

The Storage Efficiency Example

All sorts of interest in saving power and environmental footprint through the use of new storage technologies: big drives, enterprise flash, dedupe, etc.

But, in order to use these technologies effectively, you've got to understand your information to a higher degree than you might today.

You've got to understand how it's generated.  How it's used, and by who.  How its usage pattern changes over time. How long you have to keep it around.  And so on.

I have been in more than one customer discussion where they shared things like "you know, when we went and looked at this, we realized we had multiple systems generating the same kind of information, and redundant sources for the same information, and we really didn't have an information management policy, and I don't even want to talk about what we found when we went looking at how we did backups, and ..."

I don't want to hazard a guess as to the total economic impact associated with "understanding your information and how it flows better", but I'd offer that -- at least this customer -- would agree that it far outweighed any energy / footprint savings they had in mind when they started.

So, What's The Point Here?

Dick makes the point that it's not really about "green IT", it's more about "efficient IT" -- doing things efficiently, which creates all sorts of benefits, including power and environmental footprint reductions.

If we start tackling this problem from the green side, so be it.  But don't limit your benefits or impacts to only those things that can be directly tied to reduced energy consumption or a more modest environmental footprint.

Because, if you do, I think you'll miss the big payoffs.

Comments

[Berry Zito]

Add in a dose of utility computing and you might really turn up the efficiency dial...

[Dick Sullivan]

Chuck,

Thanks for the plug.

I can attest to the benefits of the TelePresence too.

First half of last year I made several trips to Santa Clara. Now I just go down the street.

Only drawback is you don't get invited to the Pub at end of day.

Dick

[Shibin Zhang]

Chuck,

I have a degree on Environmental Chemistry and a degree on Earth Sciences. This is naturally my favorite topic.

After I finish reading this weblog, I feel your conclusion is "it all about efficient IT and there is nothing we can do". Actually, it's not true.

I remember you mentioned how fast the information volume grows in another weblog. Let’s say “total information watts = watts per information unit * information volume”. If information volume grows so fast but “watts per information unit” is a constant, then “total information watts” will increase in the same speed. I would say “it’s all about reducing watts per information unit”. There are two SOLUTIONs: 1) efficient IT and 2) new technology of increasing density and throughput while reducing power consumption. I.e. if disk A is 100 watts, 1TB and 1 gbps throughput, disk B is 100 watts, 2 TB and 2gbps throughput, then we can claim that disk B is twice as green as A since one B can be configured as two logic disk A. So, there are a lot EMC can do.

I also have an idea to measure who is the good citizen in storage industry. We can periodically publish “watts shipped / TB shipped” and “watts shipped / gbps shipped” of each storage company. People can immediately identify who is green and who is trying their best to be green (via the trend of those two data). If it works, we can introduce our experience to other industries.

Shibin

[Dennis D]

I hope you would at least agree that the icing is the sweatest part of the cake in your anology above!
There's no doubt that the best way to drive change in the DC is to focus people on the wide range of efficiencies that will increase/decrease top/bottom lines. However, to me at least there's no better business decision than one that makes money for company and makes the world a little better for my kids!

[Shibin Zhang]

I can't agree. Two reasons:

1. If everybody contributes a little better, then the result is "a lot better". If everybody contributes a little worse, then the result is "a lot worse".

2. Who says that green disk couldn't be a new market opportunity? If your sales force can make it a "trend" or "fashion", it might bring a lot of money.