Dean Nelson, eBay’s vice president of Global Foundation Services, says the effort, called the Digital Service Efficiency report, “is the miles per gallon measure for technical infrastructure for eBay.” Essentially, the company has boiled its business down to a single currency — transactions (specifically URL requests) associated with users’ buying and selling on the site — and created a slew of metrics that measure how efficiently it delivers those transactions in terms of revenue, performance, cost and carbon footprint.

The project has been about 18 months in the making, Nelson told me during a recent phone call, and eBay was finally able to set a baseline measurement of its performance in 2012. Now that it knows what’s in place and how its infrastructure performs over the course of a year, the goal in 2013 is to cut its computing-related carbon usage and costs by 10 percent and increase performance in terms of transactions per kilowatt-hour by 10 percent.

In order to meet these goals, he said, every member of the technical team — from facilities managers to software engineers — has be striving toward them and also be cognizant of how turning their “knobs” will affect the other metrics eBay is measuring. “Think of it like a Rubik’s cube,” Nelson explained. “You can solve one side but screw up the rest of them.”

eBay plans to release quarterly updates on its progress along with its earnings reports, but employees will have access to down-to-the-second visibility into what’s going on. “It makes it personal for them,” Nelson said. “They can see what their efforts mean.”

52,075 servers doing a lot of work

Nelson offered some pretty compelling examples of how the Digital Service Efficiency project works in practice. If the goal is to decrease cost per transactions, data center engineers might try to minimize power usage at the facility level while server engineers might look to lower-power gear or better utilization on existing gear. They essentially reduce the denominator in that equation “and the net result is we should make more money from those transactions,” he said.

In one real-world instance, a software engineer tweaked some code that affected how much memory an application requires and the company was able to eliminate 400 servers. That cut energy usage by 1 megawatt and a $2 million savings in capital expense when the time would have come to refresh those servers.

eBay also has created a “list of fame” and a “list of shame” that highlight the 1,000 best- and worst-utilized servers within the company. “We have a hit list,” Nelson said, and it’s going to examine the bottom 20 percent to figure out why they’re as wasteful as they are.

However, he added, it’s important to remember on the server front that improving cost, performance and carbon usage doesn’t always mean buying lower-power gear. If eBay can improve the power density of its racks using technology such as liquid cooling — something its Project Mercury data center in Phoenix is pre-equipped for — it can handle more transactions on less gear. It already has some racks running at a sustained rate of 35 kilowatts and thinks it can push that up to 50 kilowatts, Nelson said.

Clean transactions with solar panels and Bloom boxes

On the carbon front, eBay has nothing but an open field in front of it thanks to some big clean-energy projects set to go live in 2013 in its new Salt Lake City, Utah, data center called Project Topaz. For starters, it’s using Bloom Energy boxes as the primary power source, which mean a slightly higher cost per transaction, but also a 13 percent reduction in carbon emissions and increased reliability (downtime costs eBay a lot of money).

Also, the company has finally cleared some regulatory hurdles to tie an on-site solar array back to the grid. Because of changes to a Utah law that eBay lobbied for, it’s about to start sourcing off-site clean energy for its data centers, as well.

“That is a corporate priority,” Nelson said. “We want to create the cleanest commerce engine on the freakin’ planet.”

Trying to change an industry

Of course, the Digital Service Efficiency methodology isn’t the only attempt by a major data center operator to show the world how efficient it is. Google publishes annual Power Utilization Efficiency (PUE) ratings for its data centers, and Facebook occasionally does as well. On Monday, Salesforce.com released a statement underscoring its commitment to sourcing renewable energy.

However, Nelson pointed out, what eBay is doing — and encouraging others to do — is more transparent in that it gives a lot more depth about operations, including the company’s server count. Even if companies don’t publish their results, tying operational efficiency to other business objectives should have a positive effect on the bottom line and the environment, regardless. Every company will have its own base currency, Nelson explained, and they’ll have to find their own metrics to measure and figure out what are the knobs that each part of the company can turn to meet goals.

“We all have the same challenges, the same things to solve for, but we have numerous ways to solve it,” Nelson said. …”[Their implementations] may change completely, but the point is the conversation is starting.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The economics of clean-data-center innovation
• How tomorrow’s mobile-centric data centers will look
• Migrating media applications to the private cloud: best practices for businesses

If eBay’s Dean Nelson has his way, that was just the beginning. His future is one of ever-greater density in data centers driving ever-greater efficiency, and he’s relying on modular data centers like the ones Dell has provided to get him there.

Sometimes the modular data centers are the standard 8-foot by 15-foot shipping containers, and sometimes they’re more unique, custom designs. But they’re always loaded to the teeth with gear. A single unit can weigh 100,000 pounds — if you drop one in, you instantly have a whole lotta computing power in a relatively small, wholly weather-proof package.

Add a thousand servers, lower the PUE

Nelson, who serves as eBay’s senior director of global foundation services (read “he runs its data center operations”), is so excited about what his team has accomplished with the new data center that he couldn’t help but take to the whiteboard in the conference room and break into a mini-lecture on how the project came to be. Many of the nitty-gritty details on how Project Mercury shaped up — including specifications on how it’s cooled and powered –  are available in a whitepaper by the Green Grid organization. But here’s the takeaway: total cost of ownership (TCO) drives everything.

If Nelson is going to buy it, it’s going to be flexible enough to change with future generations of gear and it’s going to hit the perfect blend of density, efficiency and performance. Done right, containers and modular data centers fit the bill perfectly, and Nelson will buy them from whichever vendor is best able to meet his requirements when it’s time to load up on more capacity (and when he needs more than 1,000 servers at a time). Already, he told me, “We’ve doubled our capacity and my [operational] budget has stayed flat.”

Two HP containers, and the two Dell units in one enclosure.

Among that added capacity was 50 petabytes worth of Hadoop drives split among two 1,008-node clusters. One cluster resides at the SuperNAP data center in Las Vegas, the other in an HP container atop eBay’s new data center.

Nelson has been able to maintain a steady operational budget because it’s difficult to beat what Nelson’s containers are doing in terms of efficiency. And you can’t talk about efficiency without talking about power usage effectiveness, or PUE. It’s an industry standard for determining how much energy a data center uses for tasks other than computing (e.g., cooling or lighting). The ratio is simple enough, you divide the total power usage by the power used for computing, with 1.0 being perfect. The world’s most-efficient data centers from Google, Facebook and Yahoo top out at about 1.08.

Project Mercury gets free cooling year round, even in the heat of summer. On Aug. 23, 2011 — a 119-degree day — one of eBay’s Dell units had a partial-PUE score 0f 1.044 while drawing 520 kilowatts of power. On January 17, 2012, while drawing 1 megawatt, the same unit had consistent partial PUE of 1.018 while the rest of the data center was doing between 1.26 and 1.35. Project Mercury has room for 11 modular data center units on its roof, and every one drives down the PUE of the entire facility’s PUE. Nelson realistically expects an entire facility PUE of less than 1.2.

Big thing, small package

Given all the computing power it’s capable of consuming, though, one of the most striking things about eBay’s Project Mercury data center is its size. eBay has an existing data center next door that’s 43,000 square feet and can handle 6 megawatts of power. Project Mercury has about 14,000 square feet of computing space (including the roof), but is designed to handle 12 megawatts of power (it’s currently drawing 4 megawatts). It also costs half as much as the larger space to fill with gear and to operate.

Project Mercury has a maximum power capacity of three times its current draw for two major reasons: 1). it’s not yet full (there’s still space for seven more units on the roof alone) and 2). Nelson demands higher performance with each new generation of gear he buys. He’s even pushing modularity on the first floor that looks more like a traditional data center, which means standard-size racks with ever-more power crammed into the exact same amount of space.

It’s a strategy Nelson calls “rack and roll.” Vendors with the winning bids deliver 48U racks chock full of gear that eBay only has to plug in. When it’s time to replace racks or add new ones, his team just rolls them in and out as needed.

Right now, the HPC racks that power eBay’s search engine contain 96 servers and pull 28.8 kilowatts of power. They weigh in at 3,500 pounds apiece. Its Hadoop racks contain 48 servers, each stuffed with 12 2TB drives. The first-floor space can fit 220 of these power-packed racks, while each rooftop container can fit 20.

But they keep getting more powerful. The 28.8-kilowatt racks seem downright Herculean next to their legacy 8-kilowatt neighbors on the raised floor. When future generations surpass 40 kilowatts per rack, Nelson has equipped the facility for liquid cooling direct to the chips. On the roof, Dell boosted its performance 800 kilowatts max on the second-generation modular data center up from 550 kilowatts on the first-generation model.

At webscale, (almost) everyone’s doing it

To hear Dell’s Data Center Solutions division tell it, modular data centers aren’t exclusive to eBay. Steve Cumings, the division’s executive director, told me Dell has a good business in selling modular data centers to webscale (and other hyperscale) customers. And they like them for the sames reasons eBay does: they’re super-efficient and super-dense. Given the right cooling setup, the units can sit wherever they have access to power and act like their own little data centers.

Although Cumings isn’t allowed to give the names of most DCS customers, or even the size of their deployments, he did note that Microsoft uses Dell Modular Data Centers to power Bing Maps. He also said deployments range from eBay’s scale to “many, many, many modules.” Elsewhere, everyone from the U.S. Army to Amazon is going modular.

Better, perhaps, but will it be Tron-themed?

But eBay’s Nelson isn’t about to let anyone steal his thunder, at least when it comes to driving the most bang for his buck out of modular data centers. eBay is about to break ground on a new facility called “Project Quicksilver” in Salt Lake City, Utah, that he says is even bigger and better than Project Mercury. It hasn’t yet released details on what will be inside, but the RFP calls for a per-rack performance of up to 40 kilowatts and overall facility scalability from 4 megawatts to 30 megawatts.

And Nelson — always hungry for more performance per watt — has a two-word message for server vendors that want to win eBay’s business on future buildouts: liquid cooling. For servers, it’s like the difference between fanning yourself and jumping in a pool, and he says it will be necessary as his racks achieve more power density.

“That will be the competitive advantage for a vendor,” he said “because we will buy it.”

All images courtesy of eBay.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• How tomorrow’s mobile-centric data centers will look
• Cloud computing infrastructure: 2012 and beyond
• A near-term outlook for big data

By the end of this post, you will understand why this response is annoying to me. At 15 percent, we’re in great shape: it’s plenty good for our needs. Let’s do the math and fight the snobbery.

First, let’s look at the efficiencies of other familiar uses of energy to put PV into perspective. I will act as if I’m directly addressing the PV efficiency snob, because it’s fun — and I would never be this rude in person. This may not apply to you, the reader, so please take the truculent tone in stride.

Snark Attack

So 15 percent is far too low for you? Perhaps you reason that laboratory prototypes and expensive spacecraft applications can get 40 percent-plus results, so let’s not take the plunge prematurely, given the abysmal 15 percent.

Perhaps you drive a car. Maybe you’ll stop when you realize that it converts thermal energy from burning gasoline into locomotive power at an efficiency around 15–25 percent (and this on a finite resource). We should wait for better.

Electric cars deliver battery-stored energy to the wheels at something like 85 percent efficiency. Now we’re talking. But the charging process imposes another 85 percent efficiency, and the real kicker is that the fossil fuel (or nuclear) plant supplying the electrical power is only 35 percent efficient for a net fossil-to-wheels efficiency around 25 percent (same ballpark as the gasoline car).

Hydrogen fuel cells offer no efficiency advantage in practice, achieving 20–40 percent for the round-trip hydrogen conversions, not including the efficiency of creating and delivering the electrical power to drive the formation of hydrogen.

If you’re low on energy, you might consider eating. But on second thought, our metabolic efficiency of converting chemical energy into mechanical output is similar to that of a car, so why bother? Turn up your nose.

Perhaps you are a fan of biofuels. This is perhaps the best apples-to-apples comparison to PV, being solar-driven. An Iowa corn field captures solar energy at a paltry efficiency of 1.5 percent! Okay, but we know by now that corn ethanol has a number of problems. Algae can be far more efficient, right? But even here, photosynthesis tops out at something like 5–6 percent efficiency under ideal conditions.

PV is actually rather remarkable

Considering this last point, I think it’s rather impressive that we beat biology by a factor of 3 in just a few decades of effort (biology had much longer to work on the problem). Moreover, 15 percent is perfectly adequate for our needs, as we’ll see at the end.

Qualitative assessments aside, it is rewarding to understand the origin of PV efficiency, and to appreciate that we’re not terribly far from the theoretical limit.  The point is that we shouldn’t hold out for some arbitrary efficiency before we embrace solar PV: we don’t really need the extra efficiency, and in any case, physics has something to say about how high we might expect to go.

PV basics

A photovoltaic cell is most typically a slice of crystalline silicon 200—300 μm thick. (μm = micron = micro-meter = one-millionth of a meter). The construction can either be monocrystalline — slowly grown from a large single-crystal boule, or polycrystalline, cast in an ingot and with a patchwork of crystal domains in varying orientations (translation: pretty to look at). Monocrystalline varieties have a slight advantage in efficiency: like 18 percent vs. 15 percent. The cell is doped into what we call a p-n junction, which is basically a diode. What is important here is that the junction is very near the front surface of the cell, and it is here that energy is effectively harvested.

It works like this: a photon of light comes in from the sky, penetrating some depth into the silicon. If it has enough energy (imagine a sign out front: “you must be this tall to go on this ride”), it can pop an electron out of the lattice, leaving a “hole” behind.

The big hit: spectral limit

This is all we need to know to take our first stab at an efficiency expectation. The first piece of knowledge is that photons below a certain energy cutoff called the bandgap energy (1.12 eV in silicon; corresponding to a wavelength of 1.1 μm) are not absorbed by the material: they sail right through as if going through clear glass. Second, the photons that are absorbed only need to have 1.12 eV of energy to liberate an electron out of the lattice. Any extra is wasted, popping the electron out at high speed. It rattles around the lattice, depositing its “sugar-high” as heat as it calms down.

Putting these together, we can say that even if a perfect blackbody solar spectrum is incident on the PV cell (ignoring atmospheric effects on spectrum), we lose 23 percent of the light to infrared transparency beyond 1.1 μm, plus a thermal loss that increases with increasing photon energy (shorter wavelength). The net effect is that we get to keep 44 percent for PV energy production. This ignores many other real physical limitations that we’ll address below, but it at least represents an upper limit to efficiency expectations.

We see these effects in the figure above. At 1.1 μm, the photon is well-matched to the necessary energy for liberating an electron, and we use 100 percent of its energy. As we go to shorter wavelengths, a smaller fraction of the photon energy is utilized, resulting in 33 percent of the incident energy going to waste heat.

So this most basic analysis indicates that we are doing reasonably well to capture 16 percent efficiency out of a silicon PV cell when the crudely-determined upper limit is 44 percent. This is not much different from cars or power plants, in terms of how far below the theoretical thermodynamic limit we achieve in practice.

Better than silicon?

As an aside, the bandgap energy of silicon is 1.12 eV, corresponding to a wavelength of 1.1 μm. Other semiconductor materials have different bandgap energies. Why restrict ourselves to silicon, even though it is very abundant and we benefit from substantial knowledge and experience via the computer chip industry and related enterprises? I was curious to know what would happen to our 44 percent theoretical efficiency calculation if we allow ourselves to pick any bandgap.

If we decrease the bandgap wavelength, we squander more infrared light, but use the visible-light-dominated portion of the solar spectrum more efficiently. Longer wavelength bandgaps mean more photons are available, but achieving lower efficiency at visible wavelengths. Where is the balance?

I was amazed to see silicon perched near the maximum efficiency position in this trade-off. Who knew? A more careful treatment, using the spectrum as received on the ground and effects like those explored below, finds the peak performance closer to 0.9 μm (1.38 eV), at around 34 percent.

Into the weeds: Other pernicious limitations

A word of warning: we’re about to get into the nitty-gritty here, so if you’re already feeling a little queasy, there won’t be much harm in skipping to the last paragraph in this section.

Thus far, we have only considered the effects of the input spectrum for a single-bandgap device. But other physical limitations are at play, relating to where (or if) the photon is absorbed, the path history of the generated electron and hole, surface effects, etc. Here are four effects to consider (not a complete list):

1. The expected penetration depth of the photon into the silicon depends on wavelength/energy. Photons near the bandgap can travel a very long way before being absorbed, while high-energy photons are absorbed practically at the front surface.
2. PV cells are often fabricated with a reflective back surface (also acts as the electrode), so that photons passing through the entire wafer still have a chance to be absorbed on the rebound trip. The reflective barrier also reduces heating from infrared light that otherwise would be absorbed at the back of the cell.
3. The p-n junction is at a finite depth, so the photons absorbed above this are far more vulnerable to surface loss.
4. Shorter wavelength light suffers more reflection loss at the front surface than longer wavelengths, which is what often gives a blue tint to PV cells.

Absorption length (data from this site) is shown in the logarithmic plot below. This is only the characteristic depth of absorption, but the profile at any given wavelength follows an exponentially decaying probability of absorption, set by this scale.

At a wavelength of about 0.5 μm (green light), the absorption length is about 1 μm. Shortward of this, the third effect enumerated above becomes important. Longward of a wavelength of 1.0 μm, the absorption length becomes > 200 μm, and the light often reaches the back surface, where factor 2 comes into play.

After the absorbed photon creates an electron-hole pair, the electron wanders about, bumping this way and that, with no direction in life (diffusion). If it happens to run into the p-n junction near the front surface, it gets swept across toward the front, where it joins a flock of eager electrons itching to run out into an external circuit and do some work. If it wanders off the other way (deeper into the crystal) it may never find the junction; eventually re-combining with a “hole” elsewhere — often facilitated by crystal grain boundaries and surfaces, or by defects and impurities in the crystal.

Likewise, a hole generated above the junction may wander into the junction and be pushed to the back, in an arranged marriage (recombination) with an electron returning to the back side of the cell from service in the external circuit. The junction therefore acts like a pump, pushing electrons one way and holes the other — encouraging them to participate in a flow of current through an external device.

I made a simple model to account for these effects, where the probability of being “pumped” is unity at the junction, tapering linearly to some lesser probability at the front and back surfaces (p_f and p_b, respectively). Linear makes some sense, because — as I had to prove to myself via simulation — the chance of a random walk bumping into one extreme or the other is just linearly proportional to its starting position relative to these two boundaries. If the junction always sweeps the charge, cashing in its energy, while the surface has some fixed probability of gobbling up the charge and thus forfeiting the energy, the probability relation for points between is linear. This ignores internal recombination along the way, which dis-favors long-haul paths, making the back surface “hungrier.”

Folding this effect together with the exponential absorption probability vs. depth, and allowing perfect reflection at the back, I can produce an expectation that accounts for the first three factors above. I don’t explicitly cover the front-surface reflection loss. Most new photovoltaics have an anti-reflection coating that reduces what would be a 30 percent surface reflection to just a few percent across most of the visible and near-IR band. But it gives out at the blue or near-ultraviolet end, allowing the reflection to creep back up to 30 percent. Since the PV response at the blue end is weak already due to surface losses and poor utilization of photon energy, I just absorb the extra reflection loss into the front-surface gobble probability, which is relevant primarily for short wavelengths because of their tiny penetration depths.

Okay — boy are we in the weeds here: let’s try to pull out. Putting these effects together, we get an expected efficiency of a silicon PV of 35 percent: not far off from other evaluations. Thus the real devices are in fact getting within about a factor of two of the theoretical maximum, which is better than we get in a lot of other, important domains.

The modified curves appear above. I have added a curve for the probability of conversion. Now the photons close to the bandgap mostly sail through the device, even given a second pass due to the reflection at the rear. We get high probability between 0.6–0.9 μm because the light is converted to an electron far enough from the back face, but we are not yet suffering from the front-surface inefficiencies. The probability settles out at the 50 percent level for short wavelength, which I arbitrarily assigned as the gobble-factor of both the front and back surfaces. The 35 percent result can range from 28 percent to 41 percent as I change both front and back gobble factors all the way from 0 percent to 100 percent.

In summary, we have reduced our initial 44 percent expectation to something in the neighborhood of 35 percent by considering physical processes that are practically unavoidable. We could continue this trek, accounting for all the physical phenomena that lead to 16 percent efficiency in practice, but I think I have already overdone the point: that there are really good reasons why the efficiencies will not climb to arbitrarily high values. Basic physics stands in the way, and I am left impressed with what we’ve got.

A fantastic PV tutorial

After developing the analysis above, I came across a great site explaining the fundamental physical processes involved in photovoltaics. The abundant interactive graphics are especially delightful. For the parts with which I am familiar, I find the information to be reliable and accurate. I was especially pleased to see confirmation of the collection probability scheme I implemented (you get the same linear effect in the interactive simulation if you neglect bulk recombination by increasing the diffusion length and crank up the surface recombination effect).

PV shenanigans

How is it that some lab tests or expensive spacecraft PV panels do better than the theoretical maximum calculated above? Most often, these are multi-junction devices. If we form a stack of PV junctions made from materials other than silicon, each with a different bandgap, we can more efficiently utilize the spectrum. We’d put a thin layer of material with a blue bandgap up front, followed by a green-bandgap material, and maybe silicon underneath. The longer wavelengths will sail through the first two layers and get used by the silicon. The short wavelengths, which had trouble in silicon, are more efficiently tapped by the layers in front. More of the photon energy goes into liberating the electron rather than into its velocity (heat), and more of the photons are captured.

Such devices are certainly possible to make. They are more complex, require less standard semiconductor materials, and can therefore be very expensive. For a satellite, the cost of the panels is a trivial fraction of the total cost, and launch mass means everything. So it’s worth paying a premium price to meet their power requirements in a smaller panel. For large-scale deployment, we’re likely to go cheap and low efficiency. In fact, it is more likely that a massive deployment would use thin film (amorphous silicon, e.g.) devices, which typically have efficiencies lower than 10 percent but are easier to mass-produce.

It comes down to this

This brings us to some practical matters. Returning to the PV efficiency snob, efficiency effectively maps to area. A typical location within the U.S. gets an annual average of 5 full-sun-equivalent hours per day. This means that the 1000 W/m² solar flux reaching the ground when the sun is straight overhead is effectively available for 5 hours each day. Each square meter of panel is therefore exposed to 5 kWh of solar energy per day. At 15 percent efficiency, our square meter captures and delivers 0.75 kWh of energy to the house. A typical American home uses 30 kWh of electricity per day, so we’d need 40 square meters of panels. This works out to 430 square feet, or about one sixth the typical American house’s roof (the roof area of a two-car garage). What’s the problem?

If the calculation had yielded six times the roof area, or even one times the roof area, I would see the problem. There is even a problem with one-half, or one-third, since finding a suitable portion of roof facing the equator is an issue. But at 1/6, most houses can hack it (barring shade trees, in which case it’s not better efficiency you need!). Tripling efficiency to 45 percent, if even possible, would translate to 5 percent of your roof footprint. But there’s no magic in that. We’re already to the point where it’s feasible and practical from an energetics/area point of view. Stop crinkling that nose!

In fact, we can extend this argument to the nation or world as a whole. Even at 8 percent efficiency (typical thin film multi-junction device), we could generate all primary power with a minor land footprint, as the picture below shows. Efficiency is not the bottleneck. It’s usually price. And more complex, higher purity, higher efficiency cells don’t usually lower the price.

We do not lack the area/resources on the planet to get enough energy from PV, even at half the current silicon efficiency. Other alternatives come nowhere close to being able to make this claim. As a side note, because North America uses 25 percent of the world’s energy at present, its dot may need to grow a bit, but not exorbitantly.

As reassuring as this picture is, the photovoltaic area represents more than all the paved area in the world. This troubles me. I’ve criss-crossed the country many times now, and believe me, there is a lot of pavement. The paved infrastructure reflects a tremendous investment that took decades to build. And we’re talking about asphalt and concrete here: not high-tech semiconductor. I truly have a hard time grasping the scale such a photovoltaic deployment would represent. And I’m not even addressing storage here. So while it’s physically possible, and the efficiency is sufficient to allow it, it remains a daunting challenge.

Could we even get started? Would we agree it’s the right path? Would it have much leverage against oil, given that it’s not a liquid fuel replacement? Will it always seem dreadfully expensive after being spoiled on ridiculously cheap fossil fuels? Once oil is in global terminal decline, economies will struggle to cope, and this may not seem the most opportune time to strike out on an unprecedented large-scale expenditure, whose costs and benefits will be debated hotly.

Have I ever mentioned that an easy solution is a voluntary reduction of energy demand? But this doesn’t sound like expansion/growth, so how would that idea ever gain traction?

This post originally appeared on Tom Murphy’s blog, Do the Math: Using physics and estimation to assess energy, growth, options.

Tom Murphy is an associate professor of physics at the University of California, San Diego. An amateur astronomer in high school, physics major at Georgia Tech, and Ph.D. student in physics at Caltech, Murphy has spent decades reveling in the study of astrophysics. He currently leads a project to test general relativity by bouncing laser pulses off the reflectors left on the moon by the Apollo astronauts, achieving one-millimeter-range precision. Murphy’s keen interest in energy topics began with his teaching a course on energy and the environment for nonscience majors at UCSD. Motivated by the unprecedented challenges we face, he has applied his instrumentation skills to exploring alternative energy and associated measurement schemes. Following his natural instincts to educate, Murphy is eager to get people thinking about the quantitatively convincing case that our pursuit of an ever-bigger scale of life faces gigantic challenges and carries significant risks.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• After Solyndra: analyzing the solar industry
• Flash analysis: lessons from Solyndra’s fall
• Cleantech and investment in 2013

I can think of four reasons why cloud computing is (with few exceptions) significantly more energy efficient than using in-house data centers:

1. Economies of scale. It’s cheaper for bigger cloud computing folks to make efficiency improvements because they can spread the costs over a larger server base and can afford to have more dedicated folks focused on efficiency improvements.

For example, there are usually significant fixed costs of implementing simple techniques to improve Power Usage Effectiveness (PUE), like the costs of doing an equipment inventory and assessment of data center airflow (same for implementing institutional changes like charging users per kW instead of per square foot of floor area). Whenever there are costs that are substantially fixed (i.e. only weakly related to the size of the facility), bigger operations have an advantage because they can spread the costs over more transactions, equipment, or floor area.

There’s also a substantial advantage to having “in house” expertise devoted to efficiency, instead of having staff split between different jobs. Technology changes so rapidly that it’s hard for people not devoted to efficiency to keep up as well as those that are.

2. Diversity and aggregation. More users, more diverse users, and more users in different places means computing loads are spread over the day, allowing for increased equipment utilization. Typical in house data centers have server utilizations of 5-15 percent and sometimes much less, whereas cloud facilities for major vendors are more in the 30-40 percent range.

3. Flexibility. Cloud installations use virtualization and other techniques to separate the software from the characteristics of physical servers (some call this “abstraction of physical from virtual layers”).  This sounds like a great thing for software and total costs, but why is it an energy issue?

Using this technique means that you can redesign servers to optimize them and drop certain energy costly features. For example, if software can route around physical servers that die, you no longer need to have two power supplies in each server; the death of any one particular server doesn’t matter to the delivery of IT services.

In essence, this technique redefines the concept of reliability from one that is based on the reliability of a particular piece of hardware to one that is based on the reliability of the delivery of the IT services of interest, and this is a much more sensible approach.

4. Ability to sidestep organizational issues instead of having to address them head-on (which is hard and slow). While most company in-house IT operations face the problem of a disconnect between IT departments driving server purchases, and facilities departments paying the electric bill, that problem has largely been solved for the cloud providers. They generally have one data center budget and clear responsibilities assigned to one person with decision making authority.

Economies of scale are more powerful in the cloud scenario, because you’ve gotten rid of the impediments to taking action and can allow those economies to work their magic. Finally, it’s much easier and cheaper for people stuck with the in-house organizations to use a credit card to buy cloud services instead of waiting around for their internal IT organization to get its act together.

These four big energy advantages will over time translate into more and more pressure for companies to adopt cloud services, because the economic advantages (driven by the energy advantages) are so large.  And it’s not just energy costs, it’s the capital cost of all the supporting equipment, which in a standard in-house facility can be $25,000/kW and (together with the energy costs) add up to half or more of the total costs of the facility (for details see Koomey, Jonathan G., Christian Belady, Michael Patterson, Anthony Santos, and Klaus-Dieter Lange. 2009. Assessing trends over time in performance, costs, and energy use for servers. Oakland, CA: Analytics Press.  August 17.)

Of course, there are still issues to work out. For example, people haven’t really ironed out the complexities about liability for cloud outages.  And there will always be providers who will want to have their own in-house facilities for security reasons (like big financial institutions). But even in that case, the benefits of a virtualized cloud infrastructure can be brought to the in-house facilities.

You won’t get the same diversity, but the other benefits of cloud will still be powerful. I’ve also heard of companies creating “private clouds” for use by other companies that pay in to use them on a “members-only” basis, thus dealing with the diversity and security issues. So things are evolving rapidly, but the economic benefits are so large that we’ll see a whole lot more cloud computing in coming years.

This post originally appeared on Jonathan Koomey’s blog.

Jonathan Koomey is a researcher, author, lecturer, and entrepreneur whose work spans climate solutions, critical thinking skills, and the energy and environmental effects of information technology. He’s been a Consulting Professor at Stanford University since 2004, and recently held visiting professorships at Yale (Fall 2009) and Stanford (2003–4 and Fall 2008), and worked as a researcher and scientist at Lawrence Berkeley National Laboratory (LBNL) for more than two decades

Image courtesy of Pike Research, Grigorieff Photography, GigaOM Events

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The economics of clean-data-center innovation
• A near-term outlook for big data
• NewNet Q2: Google closes the quarter with a bang

Here are a few suggestions to help you focus on completing tasks rather than managing your to-do list:

1. Pick a tool and methodology. I’m not going to dictate a specific tool, because the tool you use has to fit your needs, work well with the type of job you do and make it easy for you to get things done. The important thing is to pick a tool that works for you. For years, I used the task list in Microsoft Outlook because it was convenient, and when I switched to a Mac, I tried a bunch of different tools before eventually settling on Hiveminder. When I was consulting, I found a task list wasn’t suitable, so I used shorter lists of the next three things I needed to do. Experiment until you find a tool a and methodology that you like, and then stick with it!
2. Keep it visible. The advantage of having a task list that is integrated with your email, like Gmail Tasks or Outlook’s task list, is that every time you look at your inbox, you have your tasks right where you can see them. With my web-based task list, I keep it open on a tab all the time, and I can quickly glance at what I need to get done. If you use a standalone task manager or a simple text document, you can keep it open on your desktop and easily accessible. By keeping your tasks visible and easy to access, you’re much more likely to see them and complete your tasks.
3. Create tasks from email. One of my favorite email productivity tips is to get task items out of your email and onto your task list; having an efficient process to create tasks from email content is important. In Hiveminder, I can forward email to a special Hiveminder address with a subject line that includes commands for things like due date and priority, and the email appears as a new task in Hiveminder right away. Other task list software lets you drag and drop email onto your to-do list, or is integrated into email clients. The important thing is to have some kind of process that allows you to quickly create new tasks from email using a method that works for you.
4. Prioritize. You’ll want to use prioritization to distinguish the tasks that are critically important. I tend to use three categories — high, medium and low — which lets me quickly scan down my list for the highest priority items. While some people argue task priority should be based on the importance of the task alone, with a due date to signify urgency, I take a more pragmatic approach and set priorities based on a combination of importance and urgency.
5. Due dates. I give every task a due date. Even when I need to arbitrarily pick a date, it helps me make sure I don’t lose track of anything. I sort my task list by date and then priority, so all of my tasks for the day are at the top of my list, in a rough order of importance. This helps me stay focused on what I need to accomplish today, and it helps me get more done. For those tasks that have arbitrary due dates, I can at least look at the task on the day that I’ve marked it as due, then decide whether I should do it right away or look at it again in a few days or a few weeks.
6. Daily reality check. While you should look over your task list throughout the day, spend a minute or two every day doing a quick “reality check” on the tasks you have scheduled. First thing in the morning or at the end of the day are good times for this. What you want to focus on during the reality check phase is how much time you really have to work on your tasks and which ones are the most important. For any tasks that you know you won’t be able to do, you can bump the due date for that task out into the future or just move it way down the prioritization. Some tasks will have increased or decreased in priority relative to other tasks, so you’ll want to adjust those priorities, too. The important thing is to get rid of the clutter so that you can more easily see which tasks you need to focus on now.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Personal tools lead to practical business
• Defining work in the digital age: an analysis by GigaOM Pro
• Millennials in the enterprise, part 1: strategies for supporting the new digital workforce

Learn Keyboard Shortcuts

You can blast through your feeds with a few simple keyboard shortcuts that allow you to quickly move around without slowing down to reach for the mouse or the touchpad. It also seems that some features don’t really have a click-able counterpart, so the only way to access them is through the keyboard shortcuts. You can get a list of the available keyboard shortcuts from the Google Reader help page, but here are a few of my personal favorites that I use most frequently:

• ? – get a list of keyboard shortcuts
• j – move to the next item in the feed
• k – move to the previous item in the feed
• <space> – page down
• <enter> – open or close an item
• v – view original post
• r – refresh feed

Go Full screen

For really serious reading, you’ll want to go into full screen mode and use all of your available screen real estate for reading feeds. In full screen mode, you get a simple window showing the current feed with no additional clutter. You can navigate using the navigation shortcuts above, in addition to some shortcuts that are specific to using full screen mode:

• f – enter or exit full screen mode
• <shift>+u – show pop-up navigation menu to change feeds

Ditch the Home Page

While the home page has some interesting things like tip of the day and recently read items, if your goal is to maximize your productivity and efficiency, you should set your landing page to something else. I have all of my most important feeds in a single folder named “Critical” and I start there. You can change your start page by going to Settings -> Reader settings -> Preferences -> Start page, and select from any folder that you have created, or from a selection of other pages (All items, Starred items, etc.)

Group and Prioritize

I make extensive use of folders as a way to group and prioritize my feeds. They have become even easier to use after the recent addition of the rename folder functionality. The feeds that are most important located at the top of my navigation window. The order of the folders denotes their importance to me, but this shifts around a bit depending on my current projects. I simply drag the folders around within the subscriptions navigation pane to reorder them.

I also group things into folders based on projects or context. For example, I usually put my work-related feeds into a couple of folders grouped by topic that I can easily get through without being distracted by personal items. Keep in mind that you can also click on a folder and navigate through all of the posts within that folder across all of the feeds, so you can get through the folder more quickly than if you are navigating the individual feeds. As a result, I sometimes put critical feeds in multiple folders (critical folder and project folder) so that I can glance at it when I land on my start page of critical feeds or when I’m browsing through the project folder. Reading an item in one folder also marks it as read in any other folder, so you don’t have to worry about duplication.

Use Trends

The trends page is surprisingly interesting and useful. You can find it in the left-hand navigation pane, or with this shortcut combination: g then <shift>+t. While you can get some interesting insights into which feeds you really read, when you read them and what you clicked, the real value is in pruning your feeds. Take a look at the frequently-updated feeds section of the subscription trends; these are the high volume feeds in your reader. Now, which ones do you really still read and which ones have the zero percent read rating? You can unsubscribe from the dead weight by clicking the conveniently-located trash can, and it won’t take you long at all to reduce the clutter.

What are your favorite tips for using Google Reader?

RoxioNow powers the digital storefront for a number of retailers, including Best Buy, Blockbuster and Sears. The addition of mobile devices will give those companies even more outlets through which to sell and rent digital videos. RoxioNow is already available through a number of consumer electronics (CE) devices, including LG Blu-ray players and TiVo DVRs, and will soon have its technology embedded on TVs and Blu-ray players from Best Buy’s Insignia brand, as well as Boxee’s upcoming broadband set-top box.

The ability to stream to multiple mobile platforms also pushes forward Sonic’s plans to enable “buy once, watch anywhere” capabilities for videos that are served through its RoxioNow platform. Giving consumers the chance to pay for a piece of content on one platform and watch it on any number of devices is seen as something of a holy grail for digital video.

Sonic and Widevine have been working to enable the option as part of the Digital Entertainment Content Ecosystem (DECE), which recently unveiled its Ultraviolet DRM framework and its plan to create a digital storage locker that would keep track of a user’s digital content and the device upon which he’d be able to watch it.

Related content on GigaOM Pro: The Return of DRM (subscription required)

Now, I should admit to a dirty little secret: I’m a bit of a productivity junkie. I get an enormous amount of pleasure out of finding faster and more efficient ways to accomplish everyday tasks; I love to find better ways to gather and process information more quickly. I actively look for ways that I can streamline activities to accomplish more in less time, and I wanted to share a few of my tips.

Less Multitasking

I know, I know, you are all expert multitaskers who can accomplish more when you do multiple things at the same time. Maybe, maybe not. There have been some recent studies showing that we are more efficient when we do one thing at a time. I’ve talked before about organizing my work into chunks where I focus on specific tasks; I believe that it’s a more efficient way to work.

This tip becomes critical for corporate web workers because if you are constantly multitasking, then you seem distracted and less productive, especially when you are multitasking on conference calls. After a few times of asking people to repeat the question that you missed because you were doing something else, your boss and coworkers are likely to become suspicious about whether or not you can pay attention when working outside of the office. They don’t know if you were distracted because of email and other work or if you were distracted by the television, your kids or other home activities. We need to stay sharp and pay attention on those conference calls, so reduce the multitasking and focus on the task at hand.

News Feeds

Remote employees can’t always rely on the hallway conversations to stay caught up on industry or company news, so if we want to stay informed, we need to do some of our own legwork. Monitoring information can take a lot of your time if you aren’t efficient, but with a few tricks, you can pare it down to something more manageable. You can start by setting up a little monitoring dashboard that you can use to keep an eye on important information at a glance. Add your company blog and some news feeds that look for keywords mentioning your company or area of expertise to get started and prioritize your feeds to put the most critical ones near the top of your dashboard and searches for less important keywords near the bottom. If you want to get really efficient, you can use a tool like Yahoo Pipes to filter your information down to only the most important items. While a dashboard or feeds of keyword searches can take a little while to set up, this work will pay off over the long term. Ultimately, you want to be able to stay on top of all of the important information about your company and industry while spending very little of your precious time.

Hack Your Email

No, not that kind of hacking. I’m talking about the good hacking where you tweak your tools to better suit your needs. Let’s face it; almost everyone working in a company spends way too much time in email. The key is to be able to process your email quickly and efficiently to make sure that you are responsive without spending too much time. We all have a different way of approaching our email and different needs based on our role and the tools we are using. Here are a few of my favorite email tips that I’ve used:

• Use color to quickly pick out important email. I currently have a specific color for the people that are the most important (bosses, employees, etc.) I’ve also used colors to designate client emails when I was doing client work. You can use colors in many different ways to help you catch anything important at a glance.
• Filters are your friend. For low priority items, you can process the email immediately by moving it to a folder without spending any extra time on it. I also use tags or smart mailboxes to allow me to efficiently process groups of email while still seeing it appear in my inbox. I use this extensively for mailing lists and other lower priority email that I can scan and process quickly in batches.
• Make canned or template responses for those common questions or regular emails that you need to send. Having a template ready to go for status reports or other regular communication can save more time than you might expect over the long term. Celine wrote some handy tips on how to use canned responses in Gmail that could also be applied to other email clients.

Of course, these tips apply to everyone, and there are many more productivity tips that I could have covered (great task lists, organizational tools, etc.), but I thought they would be especially helpful for the corporate web workers.

What are your favorite productivity tips to get more done in less time?

Photo by Flickr user Ryan Ritchie used under the Creative Commons Attribution-NoDerivs 2.0 Generic license.

The list (included in full below) offers a general sense of which metro areas are really pushing for greener buildings. But the actual rank on the list shouldn’t be taken at face value, because the number of buildings in a city with superior energy efficiency doesn’t directly correlate with reductions in energy consumption. In this case, size matters.

To determine a facility’s energy performance (the first step toward an Energy Star label), the EPA compares the amount of energy used among similar types of facilities on a scale of 1-100. To qualify for the Energy Star label, a commercial building must earn a score of 75 or higher, while industrial facilities have to score in the top 25 percent (the EPA says its rating system accounts for variables such as operating conditions and regional weather data).

For example, New York City ranks as No. 10 on the EPA’s list, with only 90 buildings. But those 90 buildings include 50.4 million square feet of floor space — more than four times the total floor space of the 120 Energy Star labeled buildings in Lakeland, Fla. So Lakeland ranks higher on the list (No. 7), but its greener buildings deliver only $8.3 million in cost savings, and prevent emissions equivalent to those resulting from 6,300 homes electricity use. By comparison, New York Cities’ 90 buildings deliver $88.3 million in cost savings and prevent emissions equivalent to those from 31,000 homes.

Los Angeles, Calif. holds the top spot on the EPA list, with 293 Energy Star labeled buildings, $93.9 million in cost savings and prevention of emissions equivalent to the impact of 34,800 homes. But the 133 buildings in Houston, Tex. prevent emissions equivalent to 53,400 homes (more than any other metro area on the list), while saving only $73.9 million and earning the city slot No. 6 on the list.

That said, spurring competition among cities to boost the efficiency of commercial buildings is certainly comes as a welcome effort. According to Energy Star, the energy used in commercial buildings and manufacturing plants accounts for nearly half of the country’s power consumption and costs more than $200 billion per year — more than any other sector of the economy.

The Energy Star program is widely used to identify top energy-performing buildings, and Matthew Macko, a principal with San Francisco-based Environmental Building Strategies, has told us that part of its appeal stems from the fact that the web-based system is quick. It also helps that the label has a federal agency vouching for it that’s well known outside the building industry (the EPA).

But alternative scoring systems have begun to crop up. The American Society of Heating, Refrigerating and Air-Conditioning Engineers unveiled a program last year called Building EQ that’s meant to “expand on the type and amount of information the Energy Star program provides,” presenting buildings with a kind of report card of their energy use. A typical commercial building today would get about a C rating under Building EQ, and while an Energy Star-rated building would likely earn a B.

The basic goal of these types of programs is to highlight information about the environmental impact of a building. From there the hope is for the market to provide a premium for structures that are cleaner and more efficient, and in turn, give owners and developers an incentive to seek out the rating with greener designs.

Photo courtesy of Flickr user O b s k u r a

Related reports on GigaOM Pro (subscription required):

Green IT, Meet Green Building

Is Energy Management the Killer App for the Home Automation Market?