I get a lot of email over the course of a week. Dozens of those emails are from PR folks who are trying to get some ink or some air time for their clients. And, I'm sympathetic to them. I really am. I actually don't mind getting those emails because every now and then, there's one that really catches my fancy and makes me respond right away with a "YES!" Most get a reply, even if I'm not 100 percent tantalized by the topic. A few just get the DELETE button. Occasionally, though, there's one that crosses my screen that resonates with me, even if it's a negative resonance. This story focuses on one of those negative resonance email messages.
The email actually did resonate positively with me until I scanned one particular paragraph, which for me broke the whole deal. Here is that email in its near entirety with associated "inside my head commentary" for narrative enhancement.
Over the past year alone, there have been increasing attempts to portray the cloud as inherently “green” due, in part, to benefits of server consolidation, increased server utilization, and a reduction in on-premise, energy-intensive hardware. In fact, a Carbon Disclosure Project study claims that by 2020, large U.S. companies using the cloud could achieve energy savings of $12.3 billion. While certainly promising, it’s actually extremely difficult to measure the performance and environmental impact of the cloud due to its rapid growth, wide range of devices, and continually changing technology and business models.
[Very good information. Excellent report on Carbon emissions. You've piqued my interest.]
The fact of the matter is, however, the engine that drives the cloud is still the data center. It’s the energy efficiency of the data center, therefore, that determines whether the cloud is truly green.
[True. Now, hit me with the plan or the climax of the story. I'm really intrigued by the possibilities.]
Despite tremendous clean-energy potential, many companies are expanding their data center facilities to meet the needs of the cloud without thought of the energy sources they’re using to power them. According to a recent Greenpeace report, large companies, including Amazon, Apple and Microsoft, continue to use “dirty energy” sources, including coal, nuclear power, and diesel from large onsite generators.
[Oh, wait, did that really just say, "Greenpeace"? Yikes, that's a very negative thing to put into a pitch to me. I'm all about being clean and green but I do not wear sandals (with or without socks), I do not drive a Prius and I do not eat vegetables exclusively. Greenpeace is not something you want to say to me in a serious conversation. Incidentally, putting NRA into a pitch would get a similar reaction from me. I feel my finger sliding toward the DELETE key.]
To mitigate the negative environmental effects of “dirty energy,” increasing importance is being placed on using renewable sources of energy like geothermal, hydroelectric, wind, and solar, to power energy-intensive data center initiatives like the cloud. But, procuring such “clean energy” sources is largely dependent on where the data center is located and its proximity to such sources. And, with $450 billion a year still being spent on building out new data center space, location and energy sources become a top concern. What happens when the best “green” locations get over populated though? Can data center cloud initiatives realistically continue to use renewable sources of energy or is the rapid expansion of the cloud itself forcing companies to continue to rely on “dirty energy”?
And, then there's the final, query to interview the firm's subject matter expert on the topic. I don't want to be too harsh because the person who sent the query is very nice and no offense is meant here to her, her firm or her client but putting something as polarizing as Greenpeace into a pitch is bad. Other than that, it's a darn good topic and an intriguing pitch.
So, why didn't I go for the interview and report that information to you? It's very simple. Green energy, while interesting to discuss over cocktails, isn't practical. Look at the alternatives given in the email: Geothermal, hydroelectric, solar, wind. As the email states, location is a huge factor. The email doesn't mention seawater for cooling, which came up in our email thread is also a green alternative.
Let me explain to you what's wrong with those green alternatives.
Geothermal: Location is the biggest problem with geothermal energy. You have to have a data center in a place near geothermal vents. That usually means earthquake-prone areas. The second problem with "geo" is the expense. The expense of building and setting up removes any cost benefit you receive from going geothermal as an energy source.
Hydroelectric: Very limited locations again for this one. And, location is the biggest problem with it. The other issue is capacity. Hoover Dam is the largest hydroelectric facility in the world and its capacity is huge but limited. Hydroelectric is not a good green alternative unless you can place a data center into an existing area where this power is available. This "green" alternative also killed out four species of fish to near extinction and has disrupted the Colorado River's natural flow. So, while it's a green alternative, Greenpeace should be irritated at its impact on the environment, not using it as a selling point. I'm just saying.
Solar: Solar is a laughable alternative energy source even in locations where the sun shines 365 days per year. You can't possibly create enough electricity to run a large data center. Additionally, the expense would price you out of the data center market. However, solar is interesting as an additional or auxiliary power source. You could use solar to power the lights, security and any other non-computing energy needs from it. As a source for data center power, no, I'm sorry, it just won't work.
Wind: Unfortunately, wind power isn't viable as a data center power source either. It's expensive, it's ugly, Greenpeace should know that it kills lots of birds and energy storage is expensive and still in the research phase. It's an unpredictable source so it can never be a base energy solution. However, I think if one could couple a wind energy solution with methane, the perfect location would be Washington, D.C. -- think about it.
Seawater: This one is almost not worth discussing because of the many problems with it. For cooling, you have to pump the seawater to the data center. How close to the ocean are you going to place your data center? Seawater is highly corrosive. Coastal air is highly corrosive. Pumping seawater is expensive. Coastal locations are also very poor for data centers because of threats from tropical storms, earthquakes and tidal waves.
We all know that there are problems with so-called dirty energy sources: Oil, natural gas, coal and nuclear. However, there are some ways that data centers can thrive and be energy efficient. I'm going to tell you how. And, it's shocking to think that at this late date, it's Ken Hess who's telling the world how to build a better, more energy efficient data center. But, here goes.
Underground: Build your data centers underground. And, you don't have to go very deep to do this. Beyond two meters of depth, the earth's temperature is fairly constant regardless of location or time of year. Have you ever been in a cave? Nice and cool, isn't it? No air conditioning required. No heating required. Constant temperature achieved. The heat generated by the computing hardware can be vented to the surface for use in home heating, office heating, animal facility heating or for a number of other purposes--be creative.
Solar/Wind: On the surface of your underground data center, you can add solar and wind power sources to help move air through the data center, to circulate water for heat exchange or to power the minimal need for light and power for miscellaneous needs.
Thorium: There you have it. The answer to our future energy requirements. Thorium is a naturally-occurring, radioactive element that is abundant, requires no enrichment and is less toxic than Uranium. Thorium already comes ready to work in its natural isotopic form. There is current research on Thorium-powered cars, home reactors and power plants based on this magical element. Combine an underground data center with a self-contained Thorium reactor and you have energy to spare. If you think I'm kidding, do your own research. But Thorium is your future. It's the green alternative power source that we can live with. And, I think Greenpeace would approve of it compared to all other alternatives.
There are clean alternatives to standard data centers. We have to change the way we build them, the way we work in them and the way we power them. In my mind, we can no longer stand to waste our valuable carbon-based resources by powering cars, homes, businesses and data centers with them. We need to do it with renewable or practical alternatives.
A new reality faces us and we need to think sanely about what we do. Computing power requirements will grow exponentially and we can't allow our technology to kill us, to ruin our economies or to destroy our landscapes.
If we're to continue enjoying our technology-based world, we need to put our technology back into the caves we emerged from. By combining the future of our energy needs with the shelters of the past, we solve our problem intelligently and with good stewardship. Now, Mr. Jetson, Mr. Flintstone, go and build me a data center.
What do you think? Can we combine nuclear technology with stone age housing and find success? Talk back and let me know.
Flintstones and Jetsons are owned by Hanna-Barbera. Movie poster is from the 1987 Hanna-Barbera movie, "The Jetsons meet the Flintstones."