Green Data Center Blog

Microsoft’s Green beats Container Data Center for Media coverage

I’ve been trading some email with other data center people and they are disappointed that the Microsoft Dublin Green Data Center is getting better coverage than Microsoft’s Container Data Center.

Why? Two main points.

1) I would say the media team did a better job of having content for the European press in Dublin, than the US press for Containers.  See the EMEA web site http://www.microsoft.com/emea/presscentre/pressreleases/DublinDataCentrePR_240909.mspx  I can’t find anything on the US site for Containers.

2) Also, having two data center press events back to back almost guaranteed the second one will get less coverage.

Here is Rich Miller’s Container Data Center coverage.

Microsoft Unveils Its Container-Powered Cloud

September 30th, 2009 : Rich Miller

A look at one of the double-decker data center containers housed at the massive new Microsoft data center near Chicago. The facility includes both raised-floor space and plug-n-play bays for containers packed with servers.

As the bay door opens at Microsoft’s enormous new Chicago data center, the future backs in on a trailer. Forty-foot long containers packed with servers are unloaded with winches, and stacked two-high onto “air skates” that float on compressed air. Using the air skates, as few as four employees can move the 60-ton stack into place in Microsoft’s “container canyon” in the lower floor of the facility in Northlake, Ill.

Rich has two entries on Dublin.

http://www.datacenterknowledge.com/archives/2009/09/24/microsofts-chiller-less-data-center/

Microsoft’s Chiller-less Data Center

September 24th, 2009 : Rich Miller

An aerial view showing rooftop air handlers at Microsoft's Dublin data center, which opens today. (Image: Microsoft Corp.)

Microsoft has joined Google on the new frontier of energy efficiency – the chiller-less data center. Microsoft today announced that its huge facility in Dublin, Ireland is running without any chillers. Outside air is drawn into the facility to cool the thousands of servers powering the company’s “Live” suite of online services for users in Europe, the Middle East and Africa.

http://www.datacenterknowledge.com/archives/2009/09/28/gallery-microsofts-dublin-data-center/

Gallery: Microsoft’s Dublin Data Center

September 28th, 2009 : Rich Miller

The exterior of the new energy-efficient Microsoft data center in Dublin, Ireland.

Microsoft’s new data center in Dublin will power much of the company’s global cloud computing operation, while using far less energy and water than typically consumed in other data centers of this scale. We’ve put together a photo gallery offers a closer look at the design innovations driving its efficiency, including photos of the server room and data center interior and a diagram of the free cooling system. See our photo feature,Inside Microsoft’s Dublin Mega Data Center.

News.com has nice picture and a promise for more.

Microsoft's Chicago data center offers a merge of old and new techniques. The ground floor features sealed containers with tightly packed racks of servers, while the second floor houses more traditional server rooms.

(Credit: Microsoft)

Vote for Data Center Designer

Web Hosting Industry Review has a poll up and part of the selection is data center designer.

DATA CENTER DESIGNER

I think it bears rewarding the person that is doing a lot for data center design, taking on and completing a particularly interesting/difficult project, or introducing important new ideas - as I mentioned before, this could probably cover the green category as well.

Rob Roy
CEO and founder of Switch Communications is the formerly secretive force behind the SuperNAP in Las Vegas, a 407,000 square-foot facility promising 100 percent uptime through a tremendous investment in infrastructure and design.

Jordan Jacobs
Director of operations for PhoenixNAP, an interesting project particularly because, according to Jacobs, they set out to build a facility that could provide top end, super-redundant colocation at a price within the means of smaller customers (including hosting providers).

Olivier Sanche
The former eBay senior director of data center services and strategy was recruited last week by Apple to be its director of global data center operations. He's taking on a role that will have him involved in Apple's under-construction $1 billion data center in North Carolina, obviously a very significant facility, if just in terms of the investment, but probably in terms of how it will be employed.

Michael Manos
Senior vice president of technical services at data center company Digital Realty Trust, Manos joined the company earlier this year after previously leading the design, construction and operation of data centers at Microsoft.

Dean Nelson
founded the influential Data Center Pulse community. He very recently left Sun Microsystems to become eBay's senior director of global data center strategy, architecture and operations. He is also the co-author the Sun Blueprint, "Energy Efficient Datacenters: The Role of Modularity in Datacenter Design," and won Sun's 2008 Innovation Award.

Write-in
limit 40 chars.

Who would you vote for?

Nvidia Fermi chip powers Supercomputer for Energy and Climate Change

To research energy and climate change requires some of the faster computers.  News.com covers the release of a chip expected to be 10 times more powerful than today’s fastest supercomputer.

Supercomputer to use new Nvidia 'Fermi' chip

by Brooke Crothers

Updated at 6:40 p.m. PDT: adding additional information about Fermi chip .

Oak Ridge National Laboratory announced plans today for a new supercomputer that will use Nvidia's next-generation GPU architecture, codenamed "Fermi."

The Oak Ridge and Fermi announcements were made at Nvidia's GPU Technology developer's conference, which kicked off Wednesday in San Jose, Calif. The Fermi chip integrates three billion transistors, about three times the number of transistors in Nvidia's most powerful graphics chip now on the market. In the future, the chip will also find its way into Nvidia's GeForce product line for PCs.

Oak Ridge's supercomputer will be used for research in energy and climate change and is expected to be 10 times more powerful than today's fastest supercomputer, according to a joint statement from Oak Ridge and Nvidia. The architecture would use both graphics processing units (GPUs) from Nvida and central processing units (CPUs), according to Nvidia. Intel and Advanced Micro Devices, among others, make the CPUs.

High-end GPUs today typically contain hundreds of processing cores, allowing them to accelerate certain types of computational tasks much more efficiently, and thereby much faster, than CPUs. The new Fermi chip will have a little more than twice as many cores as current high-end Nvidia GPUs, according to Nvidia CEO Jen-Hsun Huang, jumping from 240 cores to 512 in Fermi.

Nvidia CEO Jen-Hsun Huang shows graphics card with new Fermi chip Wednesday

(Credit: Nvidia)

Fermi GPUs would enable "substantial scientific breakthroughs" that would be impossible without the new technology, said Jeff Nichols, Oak Ridge's associate lab director for Computing and Computational Sciences, in a statement.

"With the help of Nvidia technology, Oak Ridge proposes to create a computing platform that will deliver exascale computing within ten years," Nichols said. Exascale computing proposes to go beyond petaflop performance (a thousand trillion) to one million trillion calculations per second.

If you don’t have the budget of Oak Ridge National labs you can try Supermicro’s GPU solutions that I would expect to support the fermi chip.

Supermicro GPU-Optimized Supercomputing Server Solutions

The SS6016GT GPU Supercomputing Servers establish Supermicro as the true global IT hardware leader in server architecture, performance, and Green computing. Generating massively parallel processing power and unrivaled networking flexibility with two double-width GPUs or up to 5 expansion slots in a 1U form factor, the SS6016GT is performance and quality optimized for the most computationally-intensive applications. Supermicro's unique server designs with Gold Level power supplies, energy-saving motherboards and enterprise class server management optimize cooling for even the most demanding applications, providing the perfect technology platform for these impressive GPU Supercomputing Servers.
With performance bettering a small computer cluster while packing the punch (2-4 Teraflops) in a 1U, and 4U form factor, these GPU optimized solutions are ideal for Scientific computing, EDA, oil & gas exploration, military, science, and other computational intensive applications.

Medical
Imaging
Oil & Gas Exploration
Quantum
Chemistry
Financial
Simulation
3-D Rendering
& Gaming
Astrophysics

[ Press Release / Flyer 1 / Flyer 2 / In Design / White Paper]

GPU-Integrated 1U SuperServers (Graphics cards are included)

6016GT-TF-TM2
• Dual Quad/Dual-Core Intel® Xeon® processors 5500 series
• Intel® 5520 chipset with QPI up to 6.4GT/s
• Up to 96GB of Reg. ECC DDR3 DIMM 1333/1066/800 MHz SDRAM
• 2x NVIDIA Tesla M1060 GPU Cards
   PCI-E 2.0 x4 (in x16 slot - Low-Profile)
• 3x hot-swap 3.5" drive bays
• Matrox G200eW graphics controller
• 8x counter-rotating fans with optimal fan speed control
• IPMI 2.0 with virtual media over LAN and KVM-ove-LAN
• Dual LAN with Intel 82576 Gigabit Ethernet controller
• 1400W high-efficiency Gold Level (93%) Power Supply with PMbus

Microsoft’s Dublin Green Data Center Coverage

I’ve known about Microsoft’s Dublin data center, and now that there is official coverage, I can blog about it.  The Dublin area was a good opportunity for Microsoft to push for the greenest data center. 

I’ve been to Ireland many times visiting Apple and Microsoft facilities, and my name Dave Ohara so many times my name gets spelled with an apostrophe (like O’Hara).  But I am not Irish even though I have kissed the Blarney stone and bought a family crest for my Dad of O’Hara.  My Ohara surname is Japanese, but I still enjoy Ireland.  I don’t have this problem when I go to Japan. :-)

Ohara is a Japanese surname (e. g.: 小原,尾原,大原)

The Dublin Data Center manager John Dwyer is a friend  I worked before our data center days at Microsoft, and I was thinking he needs a complete coverage blog post.  So here we go.

Let’s start with John Dwyer’s ex-boss Mike Manos.

A Well Deserved Congratulations to Microsoft Dublin DC Launch

September 25, 2009 by mmanos

Today Microsoft announced the launch of their premier flagship data center facility in Dublin, Ireland.  This is a huge achievement in many ways and from many angles.    While there are those who will try and compare this facility to other ‘Chiller-less’ facilities, I can assure you this facility is unique in so many ways.   But that is a story for others to tell over time.

I wanted to personally congratulate the teams responsible for delivering this marvel and acknowledge the incredible amount of work in design, engineering, and construction to make this a reality.  To Arne, and the rest of my old team at Microsoft in DCS – Way to go!

\Mm

PS – I bet there is much crying and gnashing of teeth as the unofficial Limerick collection will now come to a close.  But here is a final one from me:

A Data Centre from a charming green field did grow,

With energy and server lights did it glow

Through the lifting morning fog,

An electrical Tir Na Nog,

To its valiant team – Way to Go!

DataCenterKnowledge has a post on Microsoft’s Dublin Data Center.

Gallery: Microsoft’s Dublin Data Center

September 28th, 2009 : Rich Miller

The exterior of the new energy-efficient Microsoft data center in Dublin, Ireland.

Microsoft’s new data center in Dublin will power much of the company’s global cloud computing operation, while using far less energy and water than typically consumed in other data centers of this scale. We’ve put together a photo gallery offers a closer look at the design innovations driving its efficiency, including photos of the server room and data center interior and a diagram of the free cooling system. See our photo feature,Inside Microsoft’s Dublin Mega Data Center.

The official press release for EMEA is here.

Microsoft Expands Cloud Computing Capabilities & Services in Europe

Microsoft’s new Dublin Data Centre to support demand for online services for business and consumers.

TextVideosImagesResources

DUBLIN, Republic of Ireland— September 24, 2009

Microsoft today announced the opening of its first ‘mega data centre’ in Europe to meet continued growth in demand for its Online, Live and Cloud services. The $500 million total investment is part of Microsoft’s long-term commitment in the region, and is a major step in realising Microsoft’s Software plus Services strategy.

Green Data Center is mentioned in this press document.

Case Studies

Greening the data centre, the Dublin Data Centre case study In Ireland, Microsoft has built its largest data centre outside the United States, using cutting-edge technology and an innovative approach to energy efficiency as part of an integrated strategy for environment sustainability.

Steve Clayton a Microsoft blogger has his post.

I’ve seen the cloud…it lives in Dublin

I’m in sunny Dublin today (yep, it’s sunny here) for the grand opening of Microsoft’s first “mega datacenter” outside of the US. What you may ask is a mega datacenter? Well basically it’s an enormous facility from we’ll deliver our cloud services to customers in Europe and beyond.

I had the chance to check the place out last month and have a full tour and it’s incredible. Okay there isn’t much to see but that’s sort of the point. It’s this big information factory that is on a scale that you’ll not see in many other places in the world and run with an astonishing level of attention to detail.

It’s also quite revolutionary and turns out to be our most efficient data center thus far. Efficiency is measured by something called PUE that essentially looks at how much power your use vs the power you consume. The ultimate PUE of course is 1.0 though the industry average is from 2-2.4. Microsoft’s data centers on average run at 1.6 PUE but this facility takes that down to 1.25 through use of some smart technology called “air”. Most datacenters rely on chillers and a lot of water to keep the facility cool – because of the climate in Dublin, we can use fine, fresh, Irish air to do the job which has significant benefits from an environmental point of view. Put simply, it saves 18 million litres of water each month.

I’ve spent some time today talking to press about this place and I left them with the title line – I’ve seen the cloud and it lives in Dublin.

 

There are data center images on the Microsoft site..

Microsoft Dublin Data Centre

Microsoft Dublin Data Centre Entrance

Embed Image

Microsoft Dublin Data Centre Aerial

Embed Image

Air-side economization architectural diagram

Embed Image

Microsoft Dublin Data Centre Server Pods

Embed Image

Microsoft Dublin Data Centre Generator Room

Embed Image

Microsoft Dublin Data Centre Rooftop Air Units

Embed Image

News spreads fast and google news has this listing.

Data Center Knowledge

Gallery: Microsoft's Dublin Data Center

Data Center Knowledge - Rich Miller - ‎20 hours ago‎

Microsoft's new data center in Dublin will power much of the company's global cloud computing operation, while using far less energy and water than ...

Microsoft Dublin Facility Uses Free Cooling Web Host Industry Review

Putting the heat on data center cooling costs Cleantech Group

all 3 news articles »

RTE.ie

Microsoft Set to Fire up Dublin Data Center

PC World - Jeremy Kirk - ‎Sep 23, 2009‎

Microsoft has opened up for business its new Dublin data center, a massive 550000-square-foot facility dedicated to serving primarily European customers. ...

Microsoft's Chiller-less Data Center Data Center Knowledge

Microsoft goes chiller-less in its new data center Consulting-Specifying Engineer

Microsoft Dublin data centre opens Irish Times

IRISHDEV.com

all 45 news articles »

I am constantly amazed at how others don’t reference a fraction of what I put above on the original sources of information. This post took me less than 1/2 hour to write, and I didn’t have to call anyone as all the information was publicly accessible and could be found through search engines.

Referencing other sources is a key way to get your blog post up in google search results, and it builds your social network.  Now I can take this post and send it on to John Dwyer, Mike Manos, Steve Clayton and many others.

Intel says Atom Make It Faster, Faster, Faster

News.com references Intel executive Sean Maloney.

Intel to rev up Atom development, executive says

by Brooke Crothers

Intel will accelerate development of the Atom processor, according to an executive, marking a different tack than the slow-but-steady strategy to date.

The Atom chip is used most prominently in Netbooks, and its hallmark has been power efficiency--not speed. But Intel will put more focus on speed, according to Sean Maloney, an Intel executive vice president.

"We'll spin Atom more frequently. Do more like a tick-tock on Atom. Make it faster, faster, faster," said Maloney in an interview at Intel Developer Forum last week.

With the growth of netbooks and competition, Intel needs to keep ahead of the performance from the competition.

What is the Evolution of the Data Center? - Business Networking at Santa Fe Institute, Nov 12 –14, 2009

NYtimes had an article on Wall Street’s math wizards forgetting a few variables – human behavior.  This article got me thinking the top issues for data center operations have a human factor.

Wall Street’s Math Wizards Forgot a Few Variables

• By STEVE LOHR

Published: September 12, 2009

IN the aftermath of the great meltdown of 2008, Wall Street’s quants have been cast as the financial engineers of profit-driven innovation run amok. They, after all, invented the exotic securities that proved so troublesome.

Enlarge This Image

James Yang

But the real failure, according to finance experts and economists, was in the quants’ mathematical models of risk that suggested the arcane stuff was safe.

The risk models proved myopic, they say, because they were too simple-minded. They focused mainly on figures like the expected returns and the default risk of financial instruments. What they didn’t sufficiently take into account was human behavior, specifically the potential for widespread panic. When lots of investors got too scared to buy or sell, markets seized up and the models failed.

That failure suggests new frontiers for financial engineering and risk management, including trying to model the mechanics of panic and the patterns of human behavior.

The interesting thing is the same companies that run huge data centers are leaders in this topic.

Much of the early work has been done tracking online behavior. The Web provides researchers with vast data sets for tracking the spread of all manner of things — news stories, ideas, videos, music, slang and popular fads — through social networks. That research has potential applications in politics, public health, online advertising and Internet commerce. And it is being done by academics and researchers at Google, Microsoft, Yahoo and Facebook.

And, there is a chance some of this is being applied to complex modeling in data centers.

One of the interesting areas I found was the topic of econophysics.

J. Doyne Farmer, a former physicist atLos Alamos National Laboratory and a founder of a quantitative trading firm, finds the behavioral research intriguing but awfully ambitious, especially to build into usable models. Instead, Mr. Farmer, a professor at the interdisciplinary Sante Fe Institute, is doing research on models of markets, institutions and their complex interactions, applying a hybrid discipline called econophysics.

To explain, Mr. Farmer points to the huge buildup of the credit-default-swap market, to a peak of $60 trillion. And in 2006, the average leverage on mortgage securities increased to 16 to 1 (it is now 1.5 to 1). Put the two together, he said, and you have a serious problem.

“You don’t need a model of human psychology to see that there was a danger of impending disaster,” Mr. Farmer observed. “But economists have failed to make models that accurately model such phenomena and adequately address their couplings.”

I’ve been sitting on this blog entry for the last week, and thanks to a social networking connection I  am attending Santa Fe Institute’s Business Networking event.

2009 Annual Business Network and Board of Trustees’ Symposium

Multi-Dimensions of Evolution

2009 is the bicentenary of the birth of Charles Darwin. This year SFI has been celebrating this event through a variety of wide-ranging lectures, symposia, and public events on the topic of evolution. The idea has been to explore the many ways in which a Darwinian and Post-Darwinian perspective on life and time have changed our science, society and metaphysics.

This November we shall continue our celebrations with a series of talks on the multi-dimensions of evolution followed by a concert performance and readings -- extending our inquiries into the world of nineteenth century, romantic exploration, and historical synthesis, spanning science and music.

Each speaker will address some unique application of evolutionary thought, describing how an evolutionary perspective has transformed our knowledge of the world. Speakers will consider how evolutionary thinking transformed their fields, and how new, post-Darwinian ideas have been evolving and generating further insights.

Organized by David Krakauer, SFI Chair of Faculty and Professor.

I am sure I’ll learn some interesting areas where the complexity of data centers can be understood with approaches used similar to the math wizards of finance.

Adding Twitter to the blogging process

I’ve been resistant to Tweet with Twitter given I already write on average two blog entries a day.  But, this week I connected my blog www.greenm3.com to my twitter feed www.twitter.com/greenm3

Here is what I did.

1) In Typepad I added my Twitter account

2) I could now publish my posts to twitter, but it was an extra step to go to TypePad posts to enable a twitter feed.  I use Windows Live Writer to write blog entries, and there is a twitter notify plug-in.

3) You can check to see if the twitter plug-in installed in the Windows Live Writer edit blog settings.

4)  When you publish a dialog box comes to confirm the publishing to Twitter.  Here is the dialog from this post.

Overall easier than I anticipated and it is just one more dialog box now that I have Windows Live Writer configured.

You can read my tweets at www.twitter.com/greenm3 which are an alternative to subscribing to my RSS feeds.

Server: The Democratization of Data

Here is a nice slide to see the growth of users, data, devices, and interactions from 2008 to 2012.  The following is a slide from Intel Developer Forum’s keynote by Paul Otellini.

Raise Server temperatures until errors, Intel demos MCA

At Intel Developer Forum Machine Check Architecture (MCA) extensions was presented, and hardware failure was simulated.

This solution is in Windows Server 2008 R2.

MCA Devices

Microsoft Windows generic hardware abstraction layers (HAL) for Intel architectures (Halx86.dll, Halapic.dll, Halmps.dll, Halia64.dll) support the Machine Check Architectures (MCA) for the Intel Pentium Pro and Itanium processors. The HAL enables Machine Check Exception (MCE) reporting for all implementation-defined errors.

For more information about the MCA-specific interface for drivers for Intel Pentium Pro and Itanium processors, see MCA Interface for Drivers.

And, Intel has contributed the code to Linux.

Intel Contributes MCA Recovery Code to the Linux* Kernel
This code will allow graceful advanced Machine Check Architecture (MCA) recovery from memory errors on systems based on the processor code-named "Nehalem-EX". configured with large amounts of memory.

and Sun has done the work to add support to Solaris on AMD and Intel.

Generic Machine Check Architecture (MCA) In Solaris

The work described below was integrated into Solaris Nevada way back in August 2007 - build 76; it has since been backported to Solaris 10. It's never too late to blog about things! Actually, I just want to separate this description from the entry that will follow - Solaris x86 xVM Fault Management.

Why Generic MCA?

In past blogs I have described x86 cpu and memory fault management feature-additions for specific processor types: AMD Opteron family 0xf revisions B-E, and AMD Opteron family 0xf revisions F and G. At the time of the first AMD work Sun was not shipping any Intel x64 systems; since then, of course, Sun has famously begun a partnership with Intel and so we needed to look at offering fault management support for our new Intel-based platforms.

With MCA, you can monitor the processor error correction to detect the relationship of rising temperatures and processor errors.

Even if thermal alarms are going off on a server if there are no errors, should you panic?

Intel Research Labs, Energy Efficiency Projects

Energy effficiency was a reoccurring theme at Intel Developer forum.  Here is a summary of the Intel Research labs  projects that are addressing power and energy use.

Wen-Hann Wang, “Innovative Research in Power and Energy Efficiency”
Director, Circuits and Systems Lab
Intel is poised to deliver dramatic improvements in the energy efficiency of computing devices. A broad set of research from Intel Labs is looking to extend beyond Intel silicon to include innovations across the platform. Wen-Hann Wang highlighted research in the key areas of circuits, architecture and platforms.
 Resilient Circuits – Under normal operating conditions, processors regularly experience dynamic variations that, left unchecked, could cause problems in operation. To protect against these potential problems, guard bands are put in place which intentionally slow the processor and also cause it to operate at higher power. Intel researchers have developed a new technology called resilient circuits that enable the system to run at faster speeds and lower power.
 These circuits detect potential problems on critical timing paths and when needed briefly re-execute at slower speeds to ensure correct results before returning to normal operation.
 Initial tests show a 21 percent throughput gain or 37 percent power reduction.
 Super Capacitors – Intel researchers have shown how super capacitors can be used to provide added power during short, peak demand cycles. Average power consumption of a laptop system is 17.5 watts, and in normal operation, intermittent power peaks can occur to more than double that. These peaks can force compromises in the choices of battery and power supply to ensure they can supply up to a steady 65 watts when needed. Wang explained how super capacitors could lead to more efficient batteries and power supplies while also enabling peak bursts of 70 watts for new features in the 2010 Core family of processors such as Turbo Mode.
 Energy Harvesting – Researchers at Intel Labs are engaged in long-term research to explore the viability and potential for harvesting alternative energy sources. Much of this potential (e.g., solar and kinetic) could be used to help extend the availability of computing.
 Low-Power Network Agent – Intel researchers have developed a low-power network agent that enables a computer or consumer device to enter a sleep state while maintaining

its network presence thereby significantly reducing its power consumption. The low-power network agent listens to network traffic and wakes the machine only for important packets.
 Platform Power Management - For meaningful improvements in energy efficiency, Intel researchers are looking beyond optimizing just a single component or device and considering platform behavior as a whole. Managing platform power well requires a broad set of changes in areas such as software, peripherals, core logic and telemetry. Intel is taking a fundamentally new approach to platform power management where the operating system provides guidance based on an understanding of the system as a whole while the hardware provides the fine grain power management across the platform to maximize efficiency. This technology will be in the ―Moorestown‖ platform which will achieve a 50x idle power reduction over the ―Menlow‖ platform.

Energy-Efficient Research has its own page on intel.com.

Home › Research › Energy-Efficiency

Energy-Efficient Systems Architecture - Intel Platform Research

As part of the Intel platform vision for architectural innovation, Intel is researching an Energy-Efficient System Architecture (EESA). EESA is a collection of technologies and architectural improvements that together will result in dramatically higher performance per watt for systems across market segments from small form factor to high performance servers. This research is concentrated on improving the power profile of Intel components, taking a systems approach to designing power efficient platforms, and collaborating with industry partners and customers to develop broader energy-efficient system solutions.

Fine Grain Power Management
back to top

At the core of EESA research, Fine-Grain Power Management is an effort to gain more precise control of power and performance across the platform. With precise power management embedded throughout the platform, performance can be maximized while power usage is precisely optimized for the workload. FGPM is the architectural foundation of Intel's research in creating dramatic improvements in energy-efficiency for systems whether small or large.

Power Delivery
back to top

EESA is researching methods to increase the efficiency of how power is delivered from the source to the place where actual computing is done, also referred to as “wall to workload.” Repeated power conversions and inefficient design in current systems result in delivery efficiency as low as 50%, meaning half of all power usage is wasted before supporting the workload. Intel is developing methods to improve power delivery to the platform and how to achieve power conversion efficiency targeting 90%.

Visibility and Control
back to top

All around a system is information that is useful in determining the best power management policies. Under Visibility and Control, Intel is researching how various environmental sensors, as well as other system information can be delivered and used to achieve the most energy-efficient platform. Sensing and analyzing both external and internal information allows the platform to make better decisions about the use of energy.

I/O Optimization
back to top

To maximize energy-efficiency across the system, the platform needs to coordinate with external devices to make sure that workload resources are used in an efficient way. Many I/O devices on a system require constant attention by the platform to function properly. The I/O Optimization research is focused on finding new ways to reduce external interfaces' constant dependency on the chipset, thereby allowing the platform to reduce power and increasing overall system performance per watt.

Power Management Policy
back to top

The components of EESA provide the tools to maximize performance per watt on the system. To make the best use of those tools, Intel is researching Power Management Policies that are designed to look at all available information, and make the best use of system resources to maximize performance when necessary, while minimizing power usage whenever possible. The Power Management Policy is the key intelligence in the EESA definition to achieve dramatic improvements in performance per watt.

Executive Blog

Join Chief Technology Officer Justin Rattner as he looks into the future at ZDNet's “Over the Horizon”blog.

Thanks to meeting some of the Intel Research engineers at IDF, I have added a stop at their labs the next time I am in Portland/Hillsboro in October.