Data center power

Eaton pushes 400/230V unit, but is it ready for (U.S.) primetime?

Last week, Eaton Corp. announced a new uninterruptible power supply (UPS) with a 400V/230V power distribution that’s uncommon in U.S. data centers.

In most data centers, power comes from the utility at 480 volts and then gets stepped down to a 208/120V distribution at the server level. With the Eaton product, the power comes from the grid at 480 volts and gets stepped down to the 400/230V distribution, which is commonly used in Europe, Asia and South America. According to an Eaton whitepaper, the 400/230V distribution is 4% more efficient than the traditional UPS that distributes at 208/120V.

“It only took us about 100 years to get to where the rest of the world has been for decades,” Peter Sacco, president of data center engineering firm PTS Data Center Solutions, wrote in an email to me. “The technical advantages of 400V are undeniable and I believe and outstanding improvement to our 208/120V standard as well as better than DC power distribution.”

But hold on.

“However, don’t buy a new electric razor just yet,” Sacco advised. “Until there is broad acceptance by the power distribution industry, don’t expect to see it widely utilized just yet.”

EPA releases documents on Energy Star for storage

The federal Environmental Protection Agency has now started down the road to developing an Energy Star spec for storage equipment. Late last month, the EPA released the first version of its Energy Star spec for servers, and it continues to work on making enterprise data centers more efficient. (At least for now, it seems that most data center managers are indifferent to Energy Star for servers.)

This week the EPA released a framework document for Energy Star storage equipment. Among other things, it looks like the spec will cover direct attached storage (DAS), network attached storage (NAS) and storage attached network (SAN), hard disk, tape, optical, solid state, hybrid storage, and bladed storage. The EPA has a dedicated Energy Star site for storage if you want to check it out.

Energy Star for servers: Unexpected impact on data center efficiency

This blog post was written by SearchDataCenter.com contributer Julius Neudorfer.

It’s out! After several years, the Environmental Protection Agency (EPA) has released the first version of the Energy Star specification for servers. Now, how fast can we adapt?

It was created through a solid collaborative effort between government and major equipment vendors. The spec addresses many areas of power usage and waste in the power supply (and redundant power supplies) for servers. Until now, there has been no standard for server power supply efficiency — the existing Energy Star program covered PCs but exempted servers. Most server manufactures have been voluntarily improving their power supply and server energy efficiency, but few published their complete specifications. The spec also addressed standby power and efficiency at less than full load. In fact, it calls for a minimum of 85% efficiency at 50% of rated load, and 82% at only 20% of rated maximum. This is an extremely important step forward, since many servers normally run with dual power supplies, each one only loaded at 20-30% of its maximum rating due to load sharing (under 2N, they normally never operate above 50%). The spec also calls for the second (redundant) power supply to have a static loss of 20 W or less. This is a major improvement to the fixed losses found in typical servers with dual power supplies.

The spec covers more than just power supplies. It even limits the idle power of hard drives to 8 W and memory to only 2 W per gigabyte. (Note: There are some limitations on this, but is part of the requirements.)

Power management is required! Moreover, the spec mandates that idle servers must draw much less power than existing servers and that power management must be enabled when shipped. Until now, most manufacturers shipped servers with the power management disabled, and most IT shops never used or chose not to enable it. In fact, the Energy Star spec calls for a base server with one CPU to draw only 55 W at idle. This is about one-third or less of the typical single CPU server, which can draw 150-200 W at idle.

However, the spec excludes blade servers! Due to the complexity of blade chassis, power supply options and server blades from different vendors, this first standard wisely decided not to further delay the release date to include blade servers. (This is still being worked on and is expected to be addressed later this year.)

I predict the specification will have unexpected effects on data center efficiency in the future. While at first blush all this should save energy in the data center, this is just the first of many mandated server and IT equipment energy-efficiency regulations that will impact the relatively flat power curve of data centers. Until recently, most servers and IT equipment drew a substantial amount of the maximum power, even while idle. As these new servers begin to replace older equipment, it will begin to be felt and seen in the IT power load, which will vary much more widely that it does today. This means the UPS and cooling loads will become much more dynamic and will require more responsive, scalable infrastructure systems to operate efficiently (on demand) at peak power and cooling loads as well as low loads.

The data center infrastructure of the future will need to be responsive to continuous and rapid changes in power demands and especially to moving and changing cooling loads, as IT equipment powers up and down in different areas of the floor. As this new paradigm evolves, The “smart” data center of the future will use advanced power management systems to interactively broker and negotiate power requests (and perhaps charges and rates) from IT equipment to the UPS, intelligent PDUs and, most importantly, intelligent cooling systems, which will need to adapt to varying heat loads while trying to operating efficiently.

Stay tuned as this specification develops. As they say in the auto business, “Your actual mileage may vary”

Syracuse University data center to be powered by microturbine generators

Syracuse University and IBM are teaming up on a new data center that will be off the utility’s power grid, instead relying on natural gas to power microturbine generators and supply both electricity and cooling to the facility.

The cost of the 6,000-square-foot data center is estimated at $12.4 million, with The New York State Energy Research and Development Authority pitching in $2 million. The server infrastructure will be a mix of IBM boxes — blades, Power-based machines, and z10 mainframes — and the data center will also use IBM’s Rear Door Heat Exchanger, which is a chilled water door on the back of server racks.

Perhaps most interesting about the project, however, will be its use of microturbine generators to power the facility (picture courtesy of Capstone, shown at right). I spoke yesterday to Roger Schmidt, a distinguished engineer at IBM who was a part of this project, about those turbines. He said the plan is for them to have 12 of the microturbine generators, each with a power capacity of 60-65 kilowatts, for a total of about 750 kilowatts maximum power to the facility. Schmidt estimated that when the facility first gets up and running, it will only need 150-200 kilowatts, and so it will be able to grow into the capacity. The data center will also be running uninterruptible power supplies (UPSes) and will be tied into the electrical grid as backups.

The microturbine generators come from Capstone MicroTurbine, one of the biggest manufacturers of these machines. Other companies that make them include Kawasaki and Solar, a division of Caterpillar. Schmidt explained how they work:

“So we bring in natural gas, which in the U.S. is pretty prevalent.  Basically we run that to a turbine, which you burn to create energy that rotates the turbine wheel. That ties into a generator that drives electricity similar to a power plant to power up the data center.”

Schmidt added that the waste heat from the turbine can then be used in two ways: to cool the data center and then for heating buildings elsewhere on campus.

How can waste heat be used for cooling? By using an adsorption chiller, the heat goes through a thermodynamic cycle, according to Schmidt, that can convert it into cooling energy. This process is typically known as cogeneration, meaning it generates both electricity and useful heat. Schmidt called it “trigeneration” because the generator not only creates electricity, but also creates useful heat that can be used in two ways — to cool the data center and warm campus buildings.

“I only know of a few other data centers using this technology of cogeneration,” Schmidt said. “It’s kind of a unique technology and really hasn’t been applied to data centers.”

Data center high density vs. low density: Is there a paradox?

Over at CIO, there is an article about a so-called data center power-density paradox. According to Michael Bullock, the CEO of a data center design consultancy called Transitional Data Services, if you don’t beware the power-density paradox, “it will ensnare you in an unappetizing manner.”

OK, so what is it? Bullock argues that as you increase the power density in your data center, “your efforts to free up space in your data center could boomerang, creating an even greater space crisis than you had before.”

Drilling down, the paradox says that as you use more dense equipment (which places greater demands on power and cooling), you will quickly reach an inversion point where more floor space is consumed by support systems than is available to your IT equipment – typically between 100 and 150 watts per square foot. This translates into greater capital and operational costs, not the reductions you were hoping to achieve.

How much space will you need?  At a power density of about 400 watts per square foot, plan to allocate about six times your usable data center space for cooling and power infrastructure.  So before you embrace high-density as a quick fix to your space problem, make sure you have adequate room to house the additional power and cooling infrastructure, sufficient raised-floor space to handle the increased airflow demands of hotter-running boxes and, of course, sufficient available power to operate the hungry systems and their support gear. If any of these resources are unavailable or inadequate, your data center will not support the increased power density. And you will have wasted your time and money.

Let’s drill down, though, for real. Let’s say you decide your data center needs to process more. As an example, let’s say you need to expand your data center so that you have 1,024 processing cores, which you calculate as 256 quad-core processors. Should you use a power-dense design, such as blade servers, or spread that processing power out amongst 1U rack servers?

Hewlett-Packard’s c7000 BladeSystem enclosure is your blade server design. In a 42U rack, you can fit four c7000s, each of which can hold 16 HP ProLiant BL2×220c G5 server blades, for a total of 64 quad-core Intel Xeon 5400 processors. That adds up to 256 quad-core processors in a single rack. Each c7000 chassis demands 6 x 2,400 watts of power, or 14,400 watts. Multiply that by four chassis and you have 57.6 kilowatts of power in a single rack holding 256 quad-core chips.

Now, let’s use a spread-out design with HP’s DL100 rack servers. A DL160 G5 rack server is 1U and holds one quad-core Intel Xeon 5400 processor. So it will take 256U, or about six 42U racks, to reach the same processing power as a single BladeSystem. Each DL160 server demands 650 watts of power, so 256 of them demands 166.4 kilowatts of power.

To sum up:

• Power-dense design: 1,024 processing cores using blade servers use up 42U of space and 57.6 kilowatts of power
• Less power-dense design: 1,024 processing cores using 1U rack servers use up 256U of space and 166.4 kilowatts of power

According to this, there is no power-density paradox. If you use power-dense equipment, you will use less space and less power.

Now, I realize that cooling a single rack of blade servers would be a ridiculously difficult chore, and would take a lot more effort than a single rack of rack servers. But that would be comparing a single blade server rack to six racks full of rack servers. It’s not an apples-to-apples comparison.

Bullock’s point is not lost. If you have a rack of 1U servers, don’t expect to be able to convert that rack to blade servers and provide the same level of power and cooling infrastructure as you presently have. It won’t happen. But that’s a comparison of more processing power to less processing power. Comparing equivalent processing power designs yield no paradox, at least on the power side of the equation.

The cooling side of the equation is a different story, and can be complicated by factors such as airside economizers, which can cool less-dense data centers but can’t cool a 57.6KW rack. So as an example, if you spread your IT equipment out enough, then maybe you could eliminate mechanical chillers altogether. That could not only cut down on space, but on cost as well (which is what matters in the end). Also, your raised floor might be able to cool six racks of 1U servers with normal CRAC units, but you might need to convert to overhead or liquid cooling to cool a single 57.6KW rack properly.

In any event, the issue is not as simple as Bullock makes it out to be. Power-dense equipment will not always lead to more data center power and cooling equipment. Oftentimes, it will lead to less when matched up against a comparable rack-server design.