Sustainable IT | Ted Samson | TAG: Servers

MORE ENTRIES

TAG: Servers

April 03, 2008 | Comments: (0)

IBM combines Systems i and p into greener Power System

IBM has made some notable green strides with its newly unveiled line of Power System servers, according to an analysis piece by Joe Clabby. The water-cooling design for example, which brings H2O directly to the microprocessor, is pretty darn innovative and energy efficient.

Overall, Clabby notes, the machine's "power, cooling, real-estate, and performance design features become even more pronounced when compared against a 64-core HP Superdome system. An 8-core Power 550 server offers 16 percent better performance than a comparably equipped HP Superdome; uses 98 percent less space; and uses 91 percent less energy."

One other green point of interest: The Power System essentially takes the place of two machines, the System i (aka AS/400), built to host OS/400, and System p, built to run on IBM AIX (Unix). The machines, released in 1999, were similar to the point that they used many of the same parts. However, IBM said at the time that it was necessary to have the two different designs -- and thus several different sets of parts -- because "[System i] had several 'unique requirements' that could only be fulfilled by using a separate chassis design and separate, specially designed microprocessor," Clabby writes.

As you may have already guessed, that's changed with the newly released Power System: It's capable of running not just one but three platforms: i (formerly i5/OS), Linux, and AIX. For IBM, it means greater savings from being able to buy more identical parts in bulk and not having to store as many different spare parts in its inventory. And from an overall green perspective, it means fewer resources are going into manufacturing and shipping some many different components.

Posted by Ted Samson on April 3, 2008 06:03 PM

February 22, 2008 | Comments: (0)

EPA grapples with Energy Star server spec

Are the servers in your datacenter energy-efficient? Well, that rather depends on what your definition of "server" is.

Bill Clintonesque as that may sound, it reflects one of the many conundrums the EPA faces as it attempts to hammer out Energy Star specifications for server hardware. The department has issued the first draft of the spec [PDF], and it's seeking input from interested stakeholders by March 14.

Interest in the topic should certainly be keen among server vendors as well as chipmakers: Given the growing trend among companies to be greener, an Energy Star stamp could prove a significant selling point for servers down the road.

The draft document reflects one of the biggest challenges the EPA faces in devising an Energy Star specification for servers: They're far more difficult to meaningfully categorize in the context of energy efficiency than end-user hardware such as PCs and laptops.

"EPA would like this specification to be as inclusive as possible but also understands that some of the more complex or niche server types and applications may not be easily addressed under this specification," the Energy Star Draft 1 document reads. "Several stakeholders have suggested that the specification focus on 'volume' servers. However, the term 'volume' is used to classify server types based more on price than function. For purposes of this specification, a computer server definition should be specific enough to clearly delineate based on intended application, hardware/software, and/or operational requirements."

That said, the draft spec proposes the following definitions:
Small Floor Standing or Rack Mounted:

2 to 4 Processor Sockets, up to 16 processor cores

2 to 16 DIMMs

1 to 6 internal disks

2-4 integrated network adaptors

4-8 option slots for adapters to expansion units or networks

1 to 5U of rack space

Medium Floor Standing or Rack Mounted:

<16 processor sockets

< 64 processor cores

< 64 DIMMs

< 16 internal disks

2-8 integrated network adapters

10-30 option slots for adapters

Typically can address upwards of 2,000 disks in expansion units

Floor-Standing and Multiplex Large Scale Servers:

< 128 processor sockets

< 1024 processor cores

< 1024 DIMMs

> 16 internal disks

2-8 integrated network adapters

More than 20 option slots for adapters

Can address upwards of > 2,000 disks

Ability to chain many systems together to present a single system environment (multiplex)

The spec also proposes defining a blade server as "a computer consisting of, at minimum, a processor, memory and hard drive that relies on certain shared resources, contained in a blade chassis, such as power supply(s), cooling, networking, system management, and storage. Blade servers are incapable of operating independent of the blade chassis."

One question the EPA seeks to answer pertains to maximum idle power requirements: How much power should an Energy-Star compliant server be permitted to consume while in idle?

"[It] is EPA's understanding that servers typically spend a significant amount of time in an idle condition. The idle state may occur at any time of the day ... when servers are consuming what could be considered high amounts of power while waiting for client requests," the draft states. "EPA believes that to the extent that the idle state for servers will persist due to numerous factors in terms of how datacenter workloads are managed, there exists the opportunity within this specification to recognize those servers that idle at lower power consumption levels. EPA is interested in receiving stakeholder feedback on the inclusion of this requirement as well as the potential use of the SPECPower test procedure and output for purposes of evaluating idle performance."

Another big question: What should the power supply efficiency requirements be? The EPA will be more specific on this key point in the second draft. For now, it says that "efficiency will be measured at several loading points. Since many of these power supplies are used in redundant configurations, EPA felt it was important to include lower loading points, such as 10% and 20% of rated output. Similarly, EPA is interested in addressing power factor at these loads in addition to the 100% load. While a power factor of 0.9 is reasonable at 100% load this level may prove more challenging at lower loading points."

Other criteria spelled out in the draft Energy Star spec for servers: "All servers must have the ability to provide real time data on AC power consumption, inlet air temperature, and processor utilization during server operation," and "qualified computer servers must come equipped with hardware power management and virtualization capabilities."

More information about the EPA's Energy Star initiative for server is available at the Energy Star Web site.

Posted by Ted Samson on February 22, 2008 12:00 PM

November 07, 2007 | Comments: (0)

Govs pledge to buy greener hardware for state governments

If you're a hardware vendor and "energy efficiency" isn't high on your product-feature list, you're risking losing some valuable customers -- perhaps as many as 50, with names such as Kansas and Minnesota.

The National Governors Association (NGA) today announced a partnership between the NGA chair's Securing a Clean Energy Future (SCEF) initiative and the Climate Savers Computing Initiative (CSCI) to spur deployment of more energy-efficient computers and servers in state offices and agencies.

Two governors, NGA Chair Gov. Tim Pawlenty of Minnesota and Kansas Gov. Kathleen Sebelius, have committed to the partnership thus far, and they will work to encourage their 48 peers to follow suit.

The move would certainly have eco-friendly benefits. Moreover, it would result fewer wasted tax dollars. Underutilized hardware can spell costly waste. For example, powering down PCs that aren't in use can result in as much as $45 in energy savings per system and $30 per monitor, per year, according to Energy Star.

"The average desktop PC currently wastes half of the power it receives," said Gov. Pawlenty in a written statement. "Having states increase the energy efficiency of their computing equipment will save consumers and taxpayers money, while reducing greenhouse gas emissions that contribute to climate change."

Under the terms of the agreement, these states have committed to reducing energy consumption from a majority of their computing equipment by:

- pledging to purchase high-efficiency computer equipment that meets or exceeds federal Energy Star ratings;

- optimizing existing computer systems by educating employees about more efficient and effective computer power-management strategies, such as better using hibernation and sleep modes;

- and ensuring in subsequent years states purchase computing equipment with increasing levels of efficiency.

Beyond preserving precious tax dollars, the initiative could spur citizens and merchants to embrace more sustainable practices themselves, observes Bill Weihl, co-chairman of the CSCI board of directors and Google's green energy czar. "[Their] actions will provide leadership for residents and local businesses looking to implement effective emission reduction policies," he said in a written statement.

CSCI, an initiative started by Google and Intel, brings together manufacturers, businesses, organizations and individuals to significantly increase the energy efficiency of computers and servers. The group is led by representatives from nine board of director companies, including Dell; Electronic Data Systems; HP; Intel; and Microsoft.

For more information about the SCEF Initiative, go to nga.org/ci. For more information about CSCI, visit climatesaverscomputing.org.

Posted by Ted Samson on November 7, 2007 10:24 AM

October 22, 2007 | Comments: (0)

Citrix unveils energy-saving PowerSmarts for Presentation Server

During peak hours, admins want their servers fully charged and running at full bore to ensure their apps are delivered without annoying service lapses. But during non-peak hours, there's no logical reason why machines need to feed on as many watts of energy.

In an effort to help cut down on that energy waste -- and the associated energy and cooling bills -- Citrix today announced a new PowerSmarts feature for its Citrix Presentation Server line, which lets admins set policies to "automatically dial server power down based on application traffic levels," according to the company.

The company claims the feature can help customers datacenter energy consumption costs by up to fifty percent, a rather impressive claim if it's accurate.

"The key to any effective response to green IT challenges is to be able to maintain, and even increase, the capacity of IT infrastructure while reducing its energy consumption," said Scott Herren, Citrix group vice president and general manager of the Citrix Application Virtualization Group. "By powering down unused Presentation Servers during off-peak usage, we are helping our customers be more green and at the same time putting money right back into their wallets."

Citrix also announced that HP is the first company with hardware to support the PowerSmarts feature in its HP ProLiant portfolio.

Notably, HP itself unveiled a power-capping feature of its own earlier this year called HP Insight Power Manager, part of the company's Systems Insight Manager (SIM) hardware management platform.

PowerSmart is compatible with all current versions and editions of Citrix Presentation Server and iLO-enabled HP servers. It will be available in December to all Citrix Presentation Server customers as a free download at www.citrix.com/cdn.

Posted by Ted Samson on October 22, 2007 02:56 PM

July 25, 2007 | Comments: (0)

EPA seeks input on Energy Star for servers

With datacenter energy consumption clearly on the U.S. government's radar, the EPA is now soliciting input from the public for developing an Energy Star standard for enterprise servers.

Energy Star energy-efficiency standards already exist for various products, including desktop hardware and peripherals, as well as major appliances such as refrigerators and washing machines. The ambitious plan to develop standards for complex servers is another indication of just how conscious the Feds have become of the environmental impact of corporate datacenters.

How this Energy Standard shapes up should be of keen interest to hardware vendors, as servers emblazoned with an Energy Star logo will prove alluring to datacenter operators who continue to struggle with the high costs of running and cooling their business-critical machinery. Fortunately, vendors and other interested parties will have a chance to weigh in.

Andrew Fanara, program manager for Energy Star product labeling, sent out a memo (PDF) late last week asking "interested stakeholders" to help shape the ES Specification Framework for Enterprise Computer Servers (PDF).

In its current state, the framework comprises three building blocks, all of which need to be further refined. They include:

definitions, which entails explicitly describing which products are covered by the specification;

eligible product categories, which identify specific product categories covered by the specification based on aforementioned definitions;

and energy efficiency criteria and test procedures for captures the energy consumed by a server "during a realistic workload."

Ideally, the EPA says it would use an industry-standard procedure for measuring energy efficiency, "but if none exists, EPA would work with stakeholders to develop an effective test procedure, according to the current framework." As it stands, EPA is monitoring work currently being done by Standard Performance Evaluation Corporation (SPEC) with the hope that it could serve as the basis for a potential Energy Star server performance metric.

As noted by my colleague Tom Yager, devising benchmarks for server energy efficiency is a complex task, one that independent testers and other organizations are struggling with. Given the varying types of server and the varying workloads they carry, it can't simply boil down to measuring, for example, how much electricity a server uses to perform X basic Web transactions. That would be like racing an unencumbered pickup truck against a hybrid at 60mph on racetrack and determining the hybrid was "better" because it got more miles per gallon. After all, the criteria would have to be different if, for example, you wanted to see which was superior, overall, for towing a heavy trailer: the hybrid or the more muscular pickup.

Interested stakeholders have until Aug. 31 to respond to the EPA's call for input on the framework. For more information, go to the Energy Star Web site.

Posted by Ted Samson on July 25, 2007 12:56 PM

July 17, 2007 | Comments: (0)

HP: Over 1B pounds recycled

Vendor aims to recycle another billion pounds of electronics and printer cartridges by 2010

HP: Over 1B pounds recycled The expression "What goes around comes around" is proving increasingly apt in the world of enterprise hardware as more companies discover the benefits of recycling retired PCs, servers, handhelds, and the like.

Indicative of the rising interest in hardware recycling, also known as IT asset recovery, HP has announced that its achieved its goal of recycling one billion pounds of hardware six months before the deadline it had set back in 2004. The company now seeks to recycle two billion pounds of gear and printer cartridges -- that is, another billion pounds -- by 2010.

"Environmental responsibility is good business," said Mark Hurd, HP chairman and CEO, in a written statement. "We've reached the tipping point where the price and performance of IT are no longer compromised by being green, but are now enhanced by it."

Pat Tiernan, VP of corporate, social, and environmental responsibility at HP, called the two billion pound goal for 2010 "the most aggressive recycling goal in the industry. "We expect to achieve it in three and a half years by expanding our convenient re-use and recycling services worldwide."

HP currently operates its program in 40 countries around the globe.

"In 2006 alone, HP recycled 164 million pounds of products globally -- the equivalent weight of more than 600 jumbo airliners and a 16 percent increase over 2005," said Tiernan.

Vendors that engage in hardware recycling refurbish and resell systems when possible ha or else mine the products for materials that can be used elsewhere. According to HP, plastics and metals it has recovered have been used to make a range of new products, including auto body parts, clothes hangers, plastic toys, fence posts, serving trays, and roof tiles.

Posted by Ted Samson on July 17, 2007 09:23 AM

June 28, 2007 | Comments: (0)

Dell chills with Emerson

Companies say bundled server and cooling products can deliver an 80 percent increase in system performance

InfoWorld: Dell chills with Emerson As datacenter managers become increasingly aware that it can cost as much to cool a server as it does to run it, hardware vendors are concocting ways to put the freeze on those high energy bills.

While some vendors, such as HP and IBM have developed their own respective cooling solutions, Dell today is announcing a different approach: The company has teamed with Emerson Network to deliver Dell-Liebert Energy Smart Solutions, which combines Dell's PowerEdge Energy Smart servers with Liebert XD and Liebert DS cooling systems.

The companies assert that the bundle can "deliver an 80 percent increase in performance and a 42 percent reduction in facility power, while maintaining high levels of business continuity and availability."

"CIOs and facilities managers are consistently faced with the challenge of increasing computing capacity to meet growing business needs while concurrently minimizing energy costs," said Rick Becker, vice president of solutions for Dell Product Group in a written statement. "By reducing both operational costs and the carbon footprint for our customers, Dell is simplifying the way businesses can take advantage of energy efficient programs and technologies without compromising IT performance."

The Liebert XD system provides high-capacity overhead or rack-adjacent cooling using a high-efficiency pumped refrigerant, which Emerson says can result in inceased datacenter cooling efficiency while letting companies use their existing datacenter infrastructure.

Liebert DS supports the XD system and uses variable capacity Digital Scroll compressors to balance cooling across the room while controlling humidity and filtering out harmful dust and particles, according to Emerson.

Dell says it's in the process of developing Energy Smart Services to complement Dell-Liebert Energy Smart Solutions. The services will help customers identify power and cooling inefficiencies, assess datacenter infrastructure and systems capacities, develop recommendations for improvements to optimize computing capacity, reclaim data center space and reduce energy usage.

The Dell-Liebert Energy Smart Solutions are available worldwide, and are being jointly marketed and sold by the two companies.

Posted by Ted Samson on June 28, 2007 08:00 AM

May 07, 2007 | Comments: (0)

OSS injects greenery into highly modular Gemini servers

Wouldn't it be nice to live in a truly plug-and-play world, where, for example, you could quickly and easily pull out the old gas-guzzling engine in your otherwise-functional automobile and swap in a hybrid engine you bought off the shelf at the local auto-parts store -- and maybe a hydrogen engine (or whatever replaces hybrid) down the road? Think of the cost and time savings, as well as the environmental benefits, of that modular model.

The automotive industry isn't there, but there's a hardware company delivering that level of flexible modularity to the server market: Open Source Systems, which today is announcing the Green Series of its 2U Gemini Server line. And in a day and age where data-center managers are keen on finding ways to make the most out of their rack space as they grow, while reducing soaring energy and cooling bills, I think this line's design holds a lot of promise.

So what makes the Gemini Green Series green? For starters, the company has added a high-efficiency power supply that runs at 93 percent efficiency. As a result, the company says the machines use between 30 and 50 percent less power than typical servers and emit up to 50 percent less heat. Chew on that for a moment: It can represent a dramatic reduction in your energy and cooling bills --and it also means you can add more servers to your data center if your problem wasn't lack of space but rather insufficient cooling.

Speaking of cooling, OSS strives to make the most of the system's 2U form factor, which houses two motherboards. The company says it's able to cool the two boards with four large 80mm fluid-bearing fans, which consume far less power (and make less noise) than the smaller five-plus wind-tunnel fans used in 1Us.

Because the system has two discrete servers, it's possible to simultaneously use an Intel processor and an AMD processors from the same chassis. That means you can choose the hardware platform that's best suited to run your services.

For example, a small company could run its Web server on an Intel processor and its database server on AMD, from one machine, and divide the duties of the 12 drives as needed. (The company notes that other 2U servers have only eight drives.)

But the real beauty in the OSS system, again, lies in its approach to modularity. "Because of its unique modular design, users can swap out major components as needs change or as technology advances so it is a truly future-proof solution," said Eren Niazi, chief executive officer and founder of Open Source Systems.

Not only can the system's major components -- the motherboard, the disk drives, power supplies, and memory -- be swapped out; they can be swapped while the system is still on the rack. The system's twelve hot-swappable disk drives can support SATA, SCSI or SAS. Mix and match as needed.

And, in a couple of years when a new crop of faster motherboards emerge with superior features that are better suited for your needs, you can easily cold-swap your existing ATX or E-ATX board simply by ejecting it via a rear-loading tray and popping in a new one. (Loosening a screw is required in this process.)

That strikes me as such an elegantly simple, straightforward, and sensible approach. From a business perspective, you're saving time and money by not having to replace the entire machine, or investing the man-hours and dollars in doing it manually.

And from an environmental standpoint, you're prolonging the lifecycle of the other perfectly good parts that you might otherwise end up shipping back to your hardware vendor for recycling when you need a system upgrade. "It's great for the environment. You're able to reuse these metal chasses without buying new servers every five year," said Niazi.

Of course, seeing is believing, and I'm looking forward to having one of the Test Center analysts give one of these babies a spin to see if they're really as good as they appear on paper. Stay tuned.

The 2U Gemini Server Green Series systems start at $10,000, but can reach $105,000, according to Niazi, depending on the tailored configuration. For more information, check out the OSS Web site at www.opensourcesystems.com.

Posted by Ted Samson on May 7, 2007 03:00 AM

May 01, 2007 | Comments: (0)

Bad corporate cultural habits lead to tolerating waste

Bad corporate cultural habits lead to tolerating waste If you've got a data center full of x86 servers -- and it's entirely likely that you do -- you should be aware that on average, you may be using less than 10% of their computing power in a 24-hour period.

Think about that for a moment: 10 percent. It's like hiring 100 employees to work for you, but having 90 or more of them sit on their tuchuses and twiddle their thumbs during any given workday. Were that the case, you'd take some presumably quick action (unless, perhaps, you'd hired them to be professional thumb-twiddlers, in which case you'd be delighted.)

That rather appalling 10-percent stat comes from a recent study by Gartner titled "Data Center Power and Cooling Scenario: Options for the Road Ahead," by the way. The author of the report, Rakesh Kumar, goes on to say that RSC/Unix are just slightly better, delivering 20% of their computer power in a 24-hour period. (He notes that "a typical mainframe environment can achieve between 70% and 80%.")

So why do organizations continue to simply purchase more x86s whenever there's a perceived need for more processing power? Because the machines are inexpensive, and organizations, much like people, seem to be creatures of habits -- many of which are bad. ("This is a result of procurement processes, project budgets and the behavior of the architectural and infrastructure teams," Kumar writes.)

It's that kind of short-sightedness that's forced so many companies into an uncomfortable position of having to scramble for space in their increasingly cramped data centers, as well as for extra scratch to pay for those increasingly growing energy bills. And the fact is, for each dollar you spend to power a server these days, you're paying another dollar to cool it.

One strategy companies can employ to kick the x86-buying addiction, and more importantly, cut energy costs and free up space in the data center, is to virtualize, the report notes (something we at InfoWorld have been professing for quite a while now).

But according to Gartner, that's but a partial solution, and some serious long-term planning is necessary if companies are going to avert the impact of the predicted power crisis that could soon hit data centers. (Check out the Gartner report for more details).

One of the real take-aways for me reading this report, though, is a lot of companies are in need of a serious wake-up call if they're willing to tolerate the kind of costly wastefulness that the Gartner report describes. Someone at your company needs to take charge at devise a strategy. Perhaps it's your CTO. Maybe you need to consider enlisting some new blood, such as a chief sustainability officer, to shake up the corporate culture. But you certainly can't afford to rest on your laurels, waiting around for some magic technology or formula to make everything better.

Posted by Ted Samson on May 1, 2007 05:30 AM

Technology White Papers

InfoWorld Technology Marketplace

» Technology White Papers Library

Technology White Papers by Topic

Technology White Papers E-mail Alert