Intel's secret weapon: Fresh air

Summary: Outside air blows hot, cold and dusty, but it cools just as well in datacenters as expensive air conditioning, according to Intel.

Fresh air could save millions in datacenter cooling costs, Intel has claimed, after a successful experiment in the New Mexico desert.

Replacing air conditioning by piping in outside air saved power costs with no appreciable increase in server failure rates, the company concluded in a research paper. Despite a lot of dust and major temperature changes--both long considered undesirable in datacenters--the equipment wasn't affected, said Intel.

"Servers... were subjected to considerable variation in temperature and humidity, as well as poor air quality; however, there was no significant increase in server failures," said the paper. "If subsequent investigation confirms these promising results, we anticipate using this approach in future, high-density datacenters."

Intel estimated an annual cost reduction of approximately $143,000 (£79,000) for a small, 500kW datacenter, based on electricity costs of eight cents per kWh. In a larger 10MW datacenter, the estimated annual cost reduction was $2.87 million.

Intel used a normal air filter that took larger particles out of the air but not fine dust. While the 32 servers and racks became coated in dust, and humidity was monitored but not controlled, the failure rate was 4.46 percent, compared with a 3.83 percent failure rate in Intel's main datacenter over the same period.

The experiment was run for 10 months, between October 2007 and August 2008. Server units with over 900 blades, used for production design, were split into two compartments. One of the compartments was air cooled, with temperatures ranging from 18 to 32°C. The other compartment was cooled using air conditioning, and used as a control.

Intel set up the experiment to challenge assumptions about optimal operating conditions in datacenters. Received wisdom has it that temperature, humidity and air quality must be strictly maintained.

However, Intel set out with the premise that, as servers are designed for optimal performance in temperatures of up to 37°C, using air cooling in desert regions could be feasible.

The experiment was run as part of the 'Intel IT's Eight-Year Data Center Efficiency Strategy' program, which aims to reduce datacenter costs.

Topics: Servers, Data Centers, Hardware, Intel, Storage

About

Tom is a technology reporter for ZDNet.com. He covers the security beat, writing about everything from hacking and cybercrime to threats and mitigation. He also focuses on open source and emerging technologies, all the while trying to cut through greenwash.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

12 comments
Log in or register to join the discussion
  • Thermal is better

    There is a better way. Use the ground to cool the system there won't be any dust or a possibility of faulty equipment with this type of air conditioning. The warmer air is piped into the ground returning as cool air. Compared to fresh air, there is no need of changing expensive air filters preventing the dust from entering new products.

    Nvidia in court for faulty chip, now Intel making contaminated processors. I'm glad that everything in my system is by AMD/ATI.
    Maarek
    • I assume you don't actually mean pumping air through the dirt...

      Unfortunately, a simple ground source heat pump might have problems keeping up with the kind of heat produced by a large server farm. As the amount of heat builds up underground, the thermal transfer rate slows down.

      Some sort of hybrid system would probably be needed to handle this kind of heat.

      "Nvidia in court for faulty chip, now Intel making contaminated processors."

      What the heck are you talking about? They were using outside air to cool a server rack, not a chip manufacturing plant. That's like saying Ford is selling contaminated trucks because they test drive them in mud.
      bmerc
  • Free air POC in Scotland

    A similar POC is going on at Sun Scotland at this moment, giving roughly the same results:
    http://wikis.sun.com/display/freeaircooling/Free+Air+Cooling+Proof+of+Concept
    petob
  • RE: Intel's secret weapon: Fresh air

    The test was about drawing outside air and expelling the hot air, not through the dirt. The humidity climate does not matter for 99.9% of locations; the target is any location that has air temps below 90f anytime; anything above 90f only requires incremental cooling. There is blogging at http://communities.intel.com/openport/blogs/datacenter/2008/09/19/free-cooling-for-data-centers-video-and-whitepaper
    Don Atwood
    • "humidity climate does not matter for 99.9% of locations"?

      How about the Midwest (or Houston) where night temps are in 60-70s and day temps are in 90-100s, with 80-99% relative humidity most of the summer? The condensation in the morning (warming outside air, cooler blade casing (probably not the blades themselves)) could possibly cause shorts. I understand that Mr. Atwood says the OEM says it is within specs to run in blah-blah % humidity, but [i]always[/i] with the caveat of [b]non-condensing[/b].

      They tested this in the desert. What is average humidity there? Were there any problems with condensation?

      Other problems are that as dust builds up, it becomes an insulator, it is especially bad for fan bearings, and dust build up on fan blades lowers their efficiency. With blade centers, you also have very confined paths for airflow, with even a small amount of blockage could take out not only that blade, but the ones next to it too. Were the air filters "standard" filters, or HEPA (costly)?

      Trust me, I would love to pipe in frest air for cooling. But, I am not sure this test was run long enough (to EQ EOL, 3-5 yrs?) to get a true set of data.
      EMonkIA
      • Interesting concept

        This realization can lead to strategic locations of datacenters in order to take advantage of the lower energy costs.

        Combining these concepts with a datacenter cooling scheme that intends to take better advantage of this type of cooling (rather than using existing strategies along with free air cooling) could realize a further reduction in energy costs.

        These tests and experiments are intended to open the eyes of those who need to know and want to know. Kudos to Intel for pushing the limits of technology.
        kc117mx
  • RE: Intel's secret weapon: Fresh air

    The fresh price of air:

    "..When a couple of electric bill
    Who were up to no good
    Startin making trouble in my datacenter
    I got in one little fight and my mom got scared
    She said 'You're movin' with your auntie and uncle in fresh Air'
    magallanes
  • 10 Months is not long enough of a test

    5 years might be more like it. Also, need to do this sort of test in a harsher environment. How about one with polution and humidity. Such as downtown Houston. It's the combination of a build up of corrosive dust and high humidity that causes problems.
    fromag82
  • we've done it for years

    Every winter, we basically force outside air into our enclosed server room (about 2 dozen servers), while capture the warm escaping air and piping into our shop. We've been doing this for over 5 years.
    shawn@...
  • A layer(s) of dust restricts heat...

    Transfer, PERIOD!

    If Intel is now manufacturing components with significantly higher heat tolerances, then they should say so. My INTEL dual Zeon 3.6 Ghz processors have a published maximum in cabinet temperature (Not to be exceeded) of 95F. Published by whom? Intel!

    A 10 degree differential for adequate cooling, depending on heat quantity produced and rate of airflow means a room temp. of 85F or less.(At a minimum)

    Layers of dust on processor heat exchangers, memory boards, hard drives, and the blades of cooling fans all combine to reduce design airflows or heat rejection efficiencies. Dust is an insulator!

    Hard drives are notoriously intolerant of heat! Raise the cabinet temperature to 100F plus, add layers of insulating dust and you will get what you sow. Penny wise and pound foolish.

    It won't happen the day you do this,(Maybe) but the destroying effects of heat will work to bring about earlier than expected component failures!

    If ALL your server/workstation components are documented from the component manufacturer to the nominal operating temps of a desert PLUS the insulating properties of dust, then go for it.Other wise, you are play Russian roulette with your computer equipment. Some people have more money (maybe not) than common sense!

    Some companies will say ANYTHING to make a buck in tough times. I have some really great real estate here in Florida I'd like to sell you! My secret weapon? Lake front property located close to the water! 50% off!
    RS9
  • RE: Intel's secret weapon: Fresh air

    Here's a good idea... not original as they've done it at the new police station here in Mildura, Australia.... Yep, it's in the Desert.

    Put at least one floor underground. Leave a space between the walls of the underground bit and the surrounding dirt, stabilise the dirt with spraycrete.

    This provides a well of cool air which can be pumped directly into the the building or minimally conditioned first. With a bit of "green" engineering like double glazing and passive thermal design, they have all but obviated the need for air conditioning. Works for me!!!
    Norm
    norm.mcmillan@...
  • RE: Intel's secret weapon: Fresh air

    try doing that in a congested urban area
    jvuntalan