Slashdot Mirror


Google Demands Higher Chip Temps From Intel

JagsLive writes "When purchasing server processors directly from Intel, Google has insisted on a guarantee that the chips can operate at temperatures five degrees centigrade higher than their standard qualification, according to a former Google employee. This allowed the search giant to maintain higher temperatures within its data centers, the ex-employee says, and save millions of dollars each year in cooling costs."

287 comments

  1. Is this possible? by CRCulver · · Score: 3, Insightful

    Wouldn't Intel run into physical limitations that simply don't allow chips to run at that low a temperature? I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round. We've seen reports of appealing places like that on Slashdot before. (Of course, that would just be a short-term fix before we move the Earth to a farther orbit around the sun to avoid suffocating in our own waste heat like the Puppeteers in Niven's Ringworld ).

    1. Re:Is this possible? by Thelasko · · Score: 5, Insightful

      Wouldn't Intel run into physical limitations

      Google is most likely getting the best chips out of Intel's standard production. It's akin to sorting out the best bananas at the grocery store. This sort of privilege happens when you buy enough products from a supplier.

      If they were demanding much more than 5 degrees then I would say they are getting custom made chips, but I don't think that's the case.

      --
      One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
    2. Re:Is this possible? by onitzuka · · Score: 4, Interesting

      WI'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      There is no reason to be surprised. It is cheaper to not move the data center to where it is colder and just make all upgraded hardware use the new chips. Google's budget calls for hardware upgrades already. Upgrading to CPUs with higher temp tolerances would mean they pay the same $X-thousand for the box they would anyway and simply turn the thermostat up.

      A net savings.

    3. Re:Is this possible? by Ngarrang · · Score: 4, Informative

      I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      Don't kid yourself. They probably have. But then, who did you get to work at what would probably be a very remote location?

      Additionally, such remote locations may suffer from not enough bandwidth and/or electricity.

      --
      Bearded Dragon
    4. Re:Is this possible? by Zocalo · · Score: 5, Interesting

      Two words: "Free Cooling"

      The greater the difference between your data centre's output air temperature and whatever passive external cooling system you are pumping it though, the more heat you can dump at zero cost. That's monetary cost as well as the cost to the environment through the energy "wasted" in HVAC systems and the like. Google has a PUE (Power Usage Effectiveness; the ratio of power input to power acutally used for powering production systems) approaching 1.2 vs typical values of around 2.5-3.0 - Microsoft is around 2.75 as I recall - so they are clearly doing something right.

      --
      UNIX? They're not even circumcised! Savages!
    5. Re:Is this possible? by iminplaya · · Score: 5, Funny

      I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      Oh yeah! And completely melt the ice cap.

      --
      What?
    6. Re:Is this possible? by try_anything · · Score: 1

      Not to mention high RTTs.

    7. Re:Is this possible? by shaka999 · · Score: 2, Interesting

      You know, they wouldn't have to go that far. I'm in Colorado, if they put a data center in one of the higher mountain towns I imagine they could significantly reduce their costs.

      I guess the other thing they could look at is using a heat exchanger and use that excess heat for hot water heating or something.

      --
      One should not theorize before one has data. -Sherlock Holmes-
    8. Re:Is this possible? by Anonymous Coward · · Score: 1, Insightful

      I'm from Colorado, too! Unfortunately, thinner air also makes cooling harder...

    9. Re:Is this possible? by jimicus · · Score: 4, Insightful

      Wouldn't Intel run into physical limitations that simply don't allow chips to run at that low a temperature? I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      Are you serious? Neither the Arctic nor the Antarctic is well known for reliable power or fast Internet connections.

    10. Re:Is this possible? by terraformer · · Score: 3, Insightful

      Two more words, uncontrolled humidity.... Yes, there are efficiency gains to be had everywhere, but none of them are free, only less costly as in bringing in moist outside air will require that air to be dehumidified, at a cost obviously. If you go with a heat exchanger, the amount of cooling decreases significantly and so you need more of them, and each one DOES require energy to operate (aka; move air/liquid/whatever through).

      --
      Who are you? The new #2 Who is #1? You are #617565. I am not a number, I am a free man! Muhahaha.
    11. Re:Is this possible? by Intron · · Score: 5, Funny

      It also takes longer to boil ramen.

      --
      Intron: the portion of DNA which expresses nothing useful.
    12. Re:Is this possible? by Dr.+Manhattan · · Score: 3, Interesting

      This happens with resistors, too. If you want one within 5% of the nominal ohmage, you pay more. If you want want one within 10%, you pay less, but you'll find that they're all either about 10% low or 10% high, with a 'notch' in the center of the distribution. Same production process used, but they skim off the 'good ones' and charge more for them.

      --
      PHEM - party like it's 1997-2003!
    13. Re:Is this possible? by thermian · · Score: 3, Insightful

      If you had say, a town sized data centre installation, it would probably have about the same effect as a smallish and partially active volcano, of which there are many in northern latitudes. Pretty much nothing apart from local effects, which is, in spite of the green crazies rantings, not too bad, not compared to the alternatives.

      What you wouldn't have is as much need for additional power to cool, which of course saves the pollution caused by its generation. You should bear in mind that the colder parts of the Earth are being far more seriously effected by polutants in the atmosphere then by anything which is just warmer than its surroundings.

      As for why I said green crazies. Well if they hadn't been so all fired determined to put governments off nuclear power, we'd have that instead of all these coal burning plants. Now we have massive pollution problems and a truly gigantic cost for building all those nuclear plants in a shorter time, instead of gradually over the last three or four decades.

      --
      A learning experience is one of those things that say, 'You know that thing you just did? Don't do that.' - D. Adams
    14. Re:Is this possible? by Colonel+Korn · · Score: 2, Interesting

      Wouldn't Intel run into physical limitations that simply don't allow chips to run at that low a temperature? I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round. We've seen reports of appealing places like that on Slashdot before. (Of course, that would just be a short-term fix before we move the Earth to a farther orbit around the sun to avoid suffocating in our own waste heat like the Puppeteers in Niven's Ringworld ).

      I doubt anything physical is being done. Intel is very conservative in setting maximum operating temperatures. They're probably just promising Google that they'll cover those operated 5 C hotter under their warranty. If anything is actually being done to the hardware it's probably just altering the temp at which throttling occurs.

      --
      "I zero-index my hamsters" - Willtor (147206)
    15. Re:Is this possible? by wvmarle · · Score: 1

      This humidity: how much is this really an issue? (genuine question). I can imagine that when you reach 100% and your equipment is cooler than ambient you get condensation issues. However here we are talking about equipment that needs cooling, i.e. is well above ambient temperatures. Condensation is therefore surely not an issue.

      If you would say "dust", that I can see as an issue as it clogs up ventilation openings and can put a nice blanket on equipment keeping it all nice and hot. Dust however is very easy to filter, particularly the larger particles of dust that are an issue in this kind of equipment.

      And finally, if a higher or varying humidity would cause more system failures, that may not be a big issue. Considering the numbers of servers Google uses this becomes really statistical, and they have already designed the whole system with failures in mind. So the failure itself is not an issue (just rip it out and replace the failed part), and cost is easily and reliably calculated using normal statistics (or own experience).

      In the case of Google, a 1-2% failure rate due to humidity issues can very well be more than offset by the savings on cooling. Cooling is very expensive after all.

    16. Re:Is this possible? by Thelasko · · Score: 4, Interesting

      Legend has it that the Celeron processor began its life as a way for Intel to make money off of the Pentiums that didn't pass quality control. If they sell the low performing processors at a discount, why shouldn't they sell the over performing ones at a premium?

      --
      One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
    17. Re:Is this possible? by Surt · · Score: 2, Interesting

      It's going to be far cheaper to build radiator fins extending into space than to move the earth's orbit, barring some innovative invention in the orbit moving department. Also, orbit moving has the downside of reducing the solar flux, which will be bad for our solar energy efforts.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    18. Re:Is this possible? by hedwards · · Score: 1

      Humidity can kill electronics. I recall some years back a guy down in Louisiana took his new camera out and about for a couple of weeks. It was a really nice camer which he'd just spent $1500 on, and within a few weeks it failed due to corrosion.

      So yes, humidity is a very serious thing and if you're not planning for it you can end up with destroyed equipment very quickly.

    19. Re:Is this possible? by Anpheus · · Score: 1

      Yeah but then they'd have to fight with Stargate Command for space.

    20. Re:Is this possible? by terraformer · · Score: 2, Insightful

      Actually it is a very good question and no one really knows the answer. Most of our IT equipment standards are derived from telecom standards from years ago. It may very well be that the tolerances are too tight and we can move away from highly controlled environments, but not enough is known about today's equipment at high or low (static problems) humidities to understand the consequences of doing so. As for dust, which is a known issue in anything with moving parts or things that don't do well with interrupted air flow, but the tighter that filter is, the air pressure needs to go up and therefore the energy needed to move air through it goes up too.

      --
      Who are you? The new #2 Who is #1? You are #617565. I am not a number, I am a free man! Muhahaha.
    21. Re:Is this possible? by aztracker1 · · Score: 2, Insightful

      Well, that's how we get the difference between a 2.4ghz and a 3.16ghz cpu as it is anyways... the higher clock ones are those that test well enough to run at that higher speed with stability. That's also why there are lower clocked CPUs that will overclock like mad (because demand is lower than yields for higher clocked parts). ... Google is probably buying the best of the best. It's not like Intel had to change a lot for that to happen. Server CPUs (XEONs, Opterons) are those that are found to stand up to higher tolerances, and usually rated under-clocked to what they could be in a desktop CPU.

      As far as I'm concerned it's interesting, but not so much news.

      --
      Michael J. Ryan - tracker1.info
    22. Re:Is this possible? by aztracker1 · · Score: 1

      Higher latitude locations would work better, you could be closer to the coastline, or a lower elevation, while still having cooler weather... Many parts of Canada, and the great lakes area would work pretty well. You don't want to move too far out thought (arctic) as this would increase latency, which needs to be a consideration as well.

      --
      Michael J. Ryan - tracker1.info
    23. Re:Is this possible? by aztracker1 · · Score: 1

      Yeah, think of the poor polar bears... Images of sad little polar bears, trapped on an ice flow come to mind... Curse you Al Gore!

      --
      Michael J. Ryan - tracker1.info
    24. Re:Is this possible? by GenP · · Score: 1

      Yet.

    25. Re:Is this possible? by rrhal · · Score: 1

      They could go to Fairbanks, AK (technically temperate sub arctic) - there's enough electricity (not as cheap as teh Hood River, OR site they currently use). It's plenty cold, there's a well educated and underutilized labor pool there, and good fiber optic networks exist between Fairbanks and Seattle/Tokyo.

      --
      All generalizations are false, including this one. Mark Twain
    26. Re:Is this possible? by TheRaven64 · · Score: 5, Informative

      It's not just the celerons, its most CPUs. A modern CPU is quite big and contains a lot of components that aren't essential to the operation of the chip. If you disable these, you have a slightly less good chip without the engineering cost of designing a entirely new die layout. AMD takes this to extremes. Their Opterons have 4 cores, three hypertransport connections and a load of cache. If there is a manufacturing flaw in the cache, they are sold as a model with less cache. If it's in the cores, then they are sold as three, two, or single core chips. If it's in the HT controllers then they only support 2- or 4-way multichip configurations. Intel's 486SX line was just a 486 (later renamed the 486DX) where the FPU didn't work.

      --
      I am TheRaven on Soylent News
    27. Re:Is this possible? by Thelasko · · Score: 1

      Google is most likely getting the best chips out of Intel's standard production.

      I should also note that Intel is guaranteeing these processors will survive at temperatures 5 degrees higher. They may not even sort through the processors they sell to Google. They are just betting that they will survive based on failure statistics they already have. They may have a higher failure rate with a 5 degree increase in temperature, but the cost of the warantee replacements is more than offset by the premium they charge Google.

      The key point is that Intel isn't custom making these processors, they are just playing games with the failure statistics.

      --
      One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
    28. Re:Is this possible? by mikael · · Score: 1

      One office block I once visited had an IBM mainframe located in the basement. One of the problems that the data center operators had, was with homeless people camping down beside the exhaust air vents. These vents generated so much heat that in Winter you could still walk around in this "bubble" of warm air in a short-sleeved shirt, while 3-4 metres away you would be frozen to your bones.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    29. Re:Is this possible? by Anonymous Coward · · Score: 0

      There is a reason CPUs, harddrives, and other electronics have a maximum altitude rating, too: as the air becomes less dense, cooling becomes less efficient even if the ambient temperature drops. This is why data centers on mountain-tops (like where large telescopes are located) are impractical.

      A sea-level Arctic data center would make more sense, but there isn't much power or communications infrastructure up there....

    30. Re:Is this possible? by Skapare · · Score: 2, Interesting

      Iceland. It has cheap geothermal energy. And that's energy that's going to heat the Earth, anyway, similar to solar. They just need some big pipes between there are North America and Europe.

      --
      now we need to go OSS in diesel cars
    31. Re:Is this possible? by iminplaya · · Score: 1

      I honestly can't believe you, and one of the mods took that seriously. I'll be sure to insert the proper link next time.

      --
      What?
    32. Re:Is this possible? by macraig · · Score: 1

      No, not possible: you forgot about that ZPM down in Antarctica. How many data centers can ya power with a ZPM? Doctor McKay probably knows... let's ask him.

    33. Re:Is this possible? by wvmarle · · Score: 1

      In that case, assuming he didn't use it under water, I'd call it a case of poor material choice by the manufacturer. Corrosion is well understood and easy enough to prevent. A camera costing that much should be corrosion resistant. Particularly as cameras are meant for outdoor use.

      Also a camera is normally at ambient temperature, maybe lower when moving from a colder to a warmer room or so. It's a low power device. Computers are normally warmer than their environment, especially servers.

    34. Re:Is this possible? by Markspark · · Score: 1

      interesting, however implausible, since they would produce low-grade heat, that would require the use of a heat-pump to reach required temperatures. This is not economic, and most paper factories (in Sweden atleast) just dump all water below 60 degrees C in to the river.

      --
      i find your lack of faith in science disturbing!
    35. Re:Is this possible? by Markspark · · Score: 1

      except that this measure would be completely pointless, since the heat lost from earth is radiated, as there is no convection/conduction in a vacuum.

      --
      i find your lack of faith in science disturbing!
    36. Re:Is this possible? by doffy · · Score: 1

      and your toilet-water spins at a lower RPM?

    37. Re:Is this possible? by andrikos · · Score: 1
      How easily do people forget the plans?

      Let's hope this time Intel's "floating" point units will not have a bug!

    38. Re:Is this possible? by Hal_Porter · · Score: 1

      Don't kid yourself. They probably have. But then, who did you get to work at what would probably be a very remote location?

      Terrorist suspects and illegal combatants.

      --
      echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
    39. Re:Is this possible? by Surt · · Score: 1

      Radiated was the point. It would be converting the radiating surface of the earth from the surface of a sphere with degree 2 to a fractal with degree up to 2.5.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    40. Re:Is this possible? by rufty_tufty · · Score: 2, Interesting

      Best isn't a term you use in test programs. At a certain temperature a chip will run at a certain speed or it won't.
      All that's going on here is that Intel will be altering its binning process to separate out the chips that are capable of running at 75 degrees from those that can't.

      Not really any different at all from the current process of binning based on speed grade. All it'll be is a different set of parameters in the test structures will cause the chip to go into a different bin.

      Now what interests me is that if google are guaranteeing to run these chips at a minimum temperature and therefore could increase the yield by accepting chips that would otherwise have been a failure because they couldn't run at a cold temperature...

      --
      "The weirdest thing about a mind, is that every answer that you find, is the basis of a brand new cliche" -
    41. Re:Is this possible? by interstellar_donkey · · Score: 3, Funny

      Right. Those polar bears, if given the chance, would rip you to shreds, eat your body and then climb into the oh-so-warm data center to hunker down for hibernation.

      My brother was a sysadmin in Alaska and a polar bear did exactly that, you insensitive clod.

      --
      The Internet is generally stupid
    42. Re:Is this possible? by NeoSkandranon · · Score: 2, Funny

      Simply bubble an offensive smelling gas into the exhaust stream (much like propane is tainted with a chemical to make leaks immediately noticeable)

      --
      If you can't see the value in jet powered ants you should turn in your nerd card. - Dunbal (464142)
    43. Re:Is this possible? by Anonymous Coward · · Score: 0

      Wouldn't Intel run into physical limitations that simply don't allow chips to run at that low a temperature? I'm surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      Are you serious? Neither the Arctic nor the Antarctic is well known for reliable power or fast Internet connections.

      Even if they had reliable power and fast net connections, putting data center there would produce enough extra heat to speed up the melting of Arctic or Antarctic ice.

    44. Re:Is this possible? by smellsofbikes · · Score: 2, Insightful

      Yeah, I am as well.
      The problem is: the adiabatic lapse rate is about 4 degrees F per thousand feet, so you go up in altitude and the air cools down, but there's less of it. Take this to an extreme and you're in outer space, where it's *very* cold (if temperature means anything at all when there's no air) but it's *incredibly* difficult to get rid of heat because there's no air.
      Ideally you'd be at sea level or below, somewhere that it's just cold.

      I recommend Iceland. They have nearly free electricity because of all their hydroelectric plants, their average temperature is pretty low, they have a good tech population, and they could really, really use the infusion of some nice Google-scented cash right about now.

      Or people could just figure out how to build cheap water-cooling and immerse the data centers. That'd solve all the problems except for the SCUBA certification you'd need to work on the equipment.

      --
      Nostalgia's not what it used to be.
    45. Re:Is this possible? by TheLink · · Score: 1

      Well it depends on how fast they want those chips to clock.

      Intel could say, 5 degrees warmer? No problem, here are your 2GHz high temp chips.

      And somehow they're the same price as the 3GHz chips that only support normal temps. Or a bit more expensive since they're "custom".

      Not sure how much CPU google needs vs I/O throughput. If they don't need as much CPU per box then this might actually be viable.

      --
    46. Re:Is this possible? by Anonymous Coward · · Score: 0

      Equipment is usually rated for the range of 10% to 95% (or 20% to 90%, etc., somewhere around there) non-condensing humidity. Not sure whether those ratings reflect actual usage tests or are pulled out of the manufacturer's ass.
      Condensation would typically only be a problem on startup of a cold system, or if condensation occurred on nearby materials and dripped onto the hot electronics.
      However, on the other end, lack of humidity can kill the electronics easily by allowing static electricity to build up and discharge. This is a real problem when bringing in cold air for "air-side economizer" free cooling.
      Therefore, there is almost always a heat exchanger between the cool outside and the cooling fluid (air, water, refrigerant, or gycol) when getting free cooling. This is somewhat more expensive and inefficient, but it's still worth it, except in very warm and humid climates.

    47. Re:Is this possible? by Markspark · · Score: 1

      out of curiosity, do you have a ball park figure of just how much of the surface would be covered by these fins to achieve this increase?

      --
      i find your lack of faith in science disturbing!
    48. Re:Is this possible? by oldhack · · Score: 1

      It also takes longer to boil ramen.

      I don't know what this guy/gal is responding to, but it's insightful all the same. Why can't you mods get it right?!

      --
      Fuck systemd. Fuck Redhat. Fuck Soylent, too. Wait, scratch the last one.
    49. Re:Is this possible? by sketerpot · · Score: 1

      So if you wanted a more precise resistance with the 10%-off resistors, you could take one that was high and another that was low and combine them in parallel? My calculations seem to say this.

    50. Re:Is this possible? by repvik · · Score: 1

      Really? I'd say he's wrong. Water boils at lower temps at higher alts, making ramenboiling quicker.

    51. Re:Is this possible? by Surt · · Score: 1

      Well, any is better than none. You just build as many as you need to get the heat dissipation rate up to the level you need.

      In the long run, I think you consider it a process of building out an array of leaves of electricity generating solar panels that also conveniently act as a heat dump into space with maximum efficiency. Of course, by the time we have the technology to take this seriously, we have other options as well which people will probably prefer. As one example, it would by then be ultra cheap to lift people into orbit along with small space craft, so we'll probably just have a diaspora to the stars before this happens. Who would choose to live on an economically impoverished earth when the resources off-earth are comparatively unlimited.
       

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    52. Re:Is this possible? by flosofl · · Score: 1

      Water boils at lower temps at higher alts, making ramenboiling quicker.

      But since it boils at a lower temp, wouldn't it then take longer to actually *cook* something? I'm anything but a competent cook (a salad with baked chicken is a stretch for me...), but my understanding was if you lower the temperature you needed to extend the time for cooking.

      Using the famous Pull-Numbers-Out-Of-My-Ass technique: if water boils at 95C at altitude and 100C at sea level, wouldn't it follow that it would take longer to cook something in boiling water at altitude?

      --
      "This calls for a very special blend of psychology and extreme violence" - Vyvyan "The Young Ones"
    53. Re:Is this possible? by mzs · · Score: 4, Informative

      What was even weirder was the i487SX. This was a i486DX with an extra pin. If you had a i486SX system and wanted a FPU you bought the i487SX and plugged it in. Then during boot the i486SX was disabled and the i487SX was used for everything.

    54. Re:Is this possible? by mzs · · Score: 1

      No. It takes less time to get the water to boil, then it needs to cook longer.

    55. Re:Is this possible? by SanityInAnarchy · · Score: 1

      There's also the altitude and the static.

      Aside from the coolness, it's not particularly healthy to hardware.

      --
      Don't thank God, thank a doctor!
    56. Re:Is this possible? by HardCase · · Score: 1

      Precisely - and since the temperature is lower, you must cook it longer.

      Taters take a lot longer too cook, too. Same problem.

    57. Re:Is this possible? by Fulcrum+of+Evil · · Score: 1

      That's why you need a pressure cooker if you live above 3000ft.

      --
      "We returned the General to El Salvador, or maybe Guatemala, it's difficult to tell from 10,000 feet"
    58. Re:Is this possible? by jank1887 · · Score: 1

      "been a failure because they couldn't run at a cold temperature."
      I have yet to see any semiconductor device that had problems running at a 'not hot' temperature. (we're not talking cold temps here. with silicon, you don't get cold temp issues until well below room temp).

    59. Re:Is this possible? by Firethorn · · Score: 2, Informative

      Did you take into account that the you'd need to double the resistance of the two resistors to get the right resistance?

      Rt = 1/(1/R1+1/R2)

      Or you could use two half value units in series.

      Either way, the price difference between 5% resisters and 10% resisters is generally NOT a factor of 2, so if you need the closer tolerance, just go with a better band. They come in 20%, 10%, 5%, 2%, and 1%.

      On the plus size, the resisters on the whole would be able to withstand almost twice as many watts(the lower resistance one would fail first).

      Hmmm...
      http://www.alliedelec.com/Resistors

      10 kohm, .25 W, 5% = $.05
      10 kohm, .25 W, 1% = $.12

      Last I looked, 20% margin resisters were hard to find. Now it looks like 10% are hard as well. Still, even at a bit over double the price at that cheap isn't going to gain you much. The extra soldering is going to be a pain.

      --
      I don't read AC A human right
    60. Re:Is this possible? by Firethorn · · Score: 1

      Now what interests me is that if google are guaranteeing to run these chips at a minimum temperature and therefore could increase the yield by accepting chips that would otherwise have been a failure because they couldn't run at a cold temperature...

      Given that CPUs are heat generating devices, and from everything I've heard will operate even in liquid nitrogen, I don't think the cold end is much of an issue.

      If your going to be running your computer outside in Antartica* during the winter, at night, all you'd have to do is put a smaller heat sink on it.

      *Don't know why, that extra 100 watts is another 100 watts to keep your living area warm.

      --
      I don't read AC A human right
    61. Re:Is this possible? by Firethorn · · Score: 1

      But it takes condensing humidity to cause corrosion - and we're unlikely to have it on active servers warmer than their surroundings.

      Wasn't there an article a couple weeks ago about a Nevada data center running with no A/C or humitidy control at all? They only experienced about 1% greater failure rates?

      A lot of the telephone tolerance stuff first came up back when a lot of the switching was [i]mechanical[/i] in nature. Even after that, you had more analogue stuff, more discrete components, etc...

      --
      I don't read AC A human right
    62. Re:Is this possible? by LaRoach · · Score: 1

      You could, but my (admittedly limited) understanding is that you would increase noise in the circuit. Do a Google search for Johnson Noise. Probably wouldn't matter much in a digital circuit but something like a decent home brew audio amp might suffer.

    63. Re:Is this possible? by Zerth · · Score: 2, Funny

      But then they'd have to find something that smells worse than homeless people, but won't get an air quality citation. Maybe set up shop next to a corn syrup plant?

    64. Re:Is this possible? by aaronbeekay · · Score: 1

      Posting to negate my fat-fingered moderation. Disregard.

    65. Re:Is this possible? by pembo13 · · Score: 1

      I'd be happy to work in the freezing area for a part pay.

      --
      "Thanks for all the money you paid to us. We've used it to buy off ISO among other things" -Microsoft
    66. Re:Is this possible? by jmichaelg · · Score: 1

      I'm surprised Google isn't using water cooling and either dumping the heat via an underground reservoir or using it to heat the company swimming pool. Water's ability to extract heat exceeds that of air by 4 to 1 so you don't need to do as much work to cool the chips.

      This guy buried a water tank to cool his Athlon 1400 which ran very hot. It was a bit extreme but it more than met his cooling requirements. A properly engineered design could do the same thing for a large datacenter. The only cost associated with cooling is overcoming the friction losses in the plumbing. Of course, if your data center is next to a river or ocean, you can just dump the heat in the river/ocean instead. San Onofre powerplant in southern california does that but before the heated water goes out to sea, they used to run it over a lobster pen. The lobsters grew to table size much faster than they do in the wild.

    67. Re:Is this possible? by lysergic.acid · · Score: 1

      i don't know about space, but shouldn't the lower air density not be as much of a factor as the colder temperatures on high elevation mountains? i mean, the thin air on Mt. Everest doesn't seem to have any difficulty sucking the heat out of people who climb it. so i'm sure there's a point at which the air becomes much too thin for efficient heat exchange, but that point is probably far above any mountain peaks on Earth.

    68. Re:Is this possible? by Anonymous Coward · · Score: 0

      cooked ramen is already very long

    69. Re:Is this possible? by SlashDev · · Score: 1

      Regardless of where you moving data centers (cold or warm locations), you still need a temperature controlled environment, in cold weather, you need to warm up the colocation. what Google wants is not to have to control it as much :)

      --

      TOP DSLR Cameras Reviews of the top DSLRs
    70. Re:Is this possible? by smellsofbikes · · Score: 1

      That's a hard question and I'd need to be smarter than I am to answer it.
      Pressure drops off exponentially with altitude. (P(x) = P(sealevel)*e^(RT/gx), as I recall, where R's the gas constant, x is elevation, g the accelleration of gravity)) So your coolling fluid volume drops off slower as you rise -- so, you lose lots of cooling power as you move from Florida to New Mexico, and less (proportionately) as you move from NM to Colorado.
      Heat transfer is, again as I recall, a function of the fourth power of the difference between the heatsource and the heatsink.
      I don't have the math background to even begin to solve that, but anything involving fourth powers will probably swamp mere exponentials -- so, yeah, probably it's better to get cooler air at higher altitude.

      --
      Nostalgia's not what it used to be.
    71. Re:Is this possible? by Fastolfe · · Score: 1

      High humidity begets condensation. You don't need to be at 100% humidity to see condensation, you just need a surface temperature lower than the temperature of the water vapor (dew point). Low humidity is bad because it doesn't carry heat away very well.

    72. Re:Is this possible? by Free+the+Cowards · · Score: 1

      Thinner air makes a huge difference. The reason people still get ass-freezingly cold on Everest is because the temperatures are ridiculously low. But take those same ridiculously low temperatures, put them at sea-level air density, and you suddenly freeze to death twice as fast or more.

      It's probably a net gain to go up a mountain, but it's not as much of a gain as you might think from simply comparing the temperatures.

      --
      If you mod me Overrated, you are admitting that you have no penis.
    73. Re:Is this possible? by spacefight · · Score: 1

      And they need some new business models anyway after the recent troubles over there... RTT still sucks though.

    74. Re:Is this possible? by Phroggy · · Score: 3, Interesting

      You are correct, or it may not cook correctly at all at the lower temperature even with more time. For example, if you want to boil rice in Guatemala City, you have to sautee it first.

      Damn Slashdot's broken Unicode support. There should be an accent in there.

      --
      $x='S24;r)>63/* h@<5+oZ)32"5cz';$me='phroggy'x$];
      $x=~y+ -xz+\0-Tx+;print$_^chop$me for split'',$x;
    75. Re:Is this possible? by jonaskoelker · · Score: 4, Funny

      Intel's 486SX line was just a 486 (later renamed the 486DX) where the FPU didn't work.

      Mod parent +5.0000000018623, Informative

    76. Re:Is this possible? by blueZ3 · · Score: 1

      Antarctica is a desert.

      No really. Look it up.

      --
      Interested in a Flash-based MAME front end? Visit mame.danzbb.com
    77. Re:Is this possible? by blueZ3 · · Score: 1

      Plu, right now you could buy the whole place for a song.

      --
      Interested in a Flash-based MAME front end? Visit mame.danzbb.com
    78. Re:Is this possible? by treeves · · Score: 1

      Most chemical reactions and cooking are faster at higher temperatures, so GP was right as long you allow that ramen needs "cooking" (arguable, I admit).

      --
      ...the future crusty old bastards are already drinking the Kool-Aid.
    79. Re:Is this possible? by Muhammar · · Score: 1

      Iceland: Insanely cheap electricity, lots of free space inland, highly educated population currently looking for a new employer, a favorable currency exchange rate...

      --
      I doubt that we will ever figure out - and I suspect that even if we did figure out we couldn't do much about it
    80. Re:Is this possible? by Alpha830RulZ · · Score: 1

      Yes. Check any cake mix box.

      --
      I was taught to respect my elders. The trouble is, it's getting harder and harder to find some.
    81. Re:Is this possible? by hkultala · · Score: 1

      The legend is wrong.

      The original celeron was a klamath core Pentium II without the external L2 cache chips. They were fully functional cores, but packaging them alone and not spending anything on those L2 chips saved some money.

      The second version of Celeron was mencodino, which was a chip purposely designed as celeron, it was not sold at all with pentium name. It was a klamath + 128kB L2 cache on-die.

      The third version of celeron, "coppermine-128" was then pentium III with hald the cache disabled.

    82. Re:Is this possible? by CAIMLAS · · Score: 1

      What astounds me is why companies like Google aren't moving data centers to places like rural Minnesota, Michigan, or even somewhere like South Dakota. These places are colder year-round than current data center locations by quite a bit, and the winters are substantially colder. It's likely they'd not even have a cooling bill at all during the winter (and, in fact, would just duct the heat into the offices/run AC into the offices).

      Personally, I think somewhere like South Dakota would be an awesome place for a corporate data center. If a state like Texas appeals to companies like RackSpace due to its "right to work' status, they'd love a state like South Dakota (which has an even lower cost of living and per capita income with the

      Yes, most of South Dakota is pretty desolate; however, the Black Hills region is very beautiful and still quite undeveloped. There is a lot of open space, land is cheap, and the cost of living is substantially lower than the "metro areas". For a small city like Rapid City, there is a fair amount of culture and life is much more pleasant than in an urban area, or even sprawling suburbia (because RC isn't capable of supporting such a thing due to the topography).

      There are some companies which are moving to smaller towns (Northrop just put a facility in locally) to save money; it just surprises me that companies like Google, which have a geographically distributed clientel and are only dependent on a proper fiber run, aren't moving to the smaller towns. They'd get cheaper facilities, cheaper employees who (most likely) want to settle down/not jump jobs, and a more distributed network (making growth and recession within the organization a more organic, less chaotic process).

      Anyway, just thought I'd plug my town for this kind of development; I like it here and would like there to be more local opportunities. In the case of Google, it seems to me that simply requiring for processors to be able to handle a higher temperature is thinking inside the box - ie, fairly increatively. Something like this should be only one of many steps taken to cut overhead and improve the product.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    83. Re:Is this possible? by CAIMLAS · · Score: 1

      I'm sorry, we're not communicating. Let's take a step back.

      We aren't heating data centers (intentionally), therefore humidity is not an issue. Outside air - the cold variety - is less moist than the warm variety. This is due to water moisture (both vapor and liquid) freezing into what is called "ice". This ice does not travel through the air freely, and falls to the ground. Additionally, cold air has less buoyancy for moisture in general, and water moisture does not readily evaporate in cold air.

      This is significant when your data center is located in a semi-arid climate which has a winter temperature around 0F. There is no moisture in the air, and, in fact, many people actually run humidifiers to cut down on static electricity and keep the air moist enough to not be uncomfortable to breath when it is at a habitable temperature.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    84. Re:Is this possible? by terraformer · · Score: 1

      And this is somehow different than what I said how? You used an example of a 0F location in semi-arid conditions. Living in a temperate climate, I am always thinking of ways to remove humidity. Your example assumes no spring, summer or fall and may as well be inside the arctic circle. You still made my point, You need to run some sort of device to modify the humidity levels.

      --
      Who are you? The new #2 Who is #1? You are #617565. I am not a number, I am a free man! Muhahaha.
    85. Re:Is this possible? by Agripa · · Score: 1

      Unless you are doing something weird, the Johnson noise will be proportional to the Thevenin resistance no matter what series-parallel combination you use. Four 10 ohm resistors in series-parallel have the same Johnson noise as a single 10 ohm resistor.

    86. Re:Is this possible? by Agripa · · Score: 1

      I have definitely encountered this but reputable resistor manufacturers do not engage in such shenanigans.

    87. Re:Is this possible? by GWBasic · · Score: 1

      surprised Google isn't considering moving some of its data centres to Arctic locations where you get cool temperatures year-round.

      My guess is that it's probably rather expensive to run fiber and power up there.

    88. Re:Is this possible? by smithmc · · Score: 1

      Iceland. It has cheap geothermal energy. And that's energy that's going to heat the Earth, anyway, similar to solar. They just need some big pipes between there are North America and Europe.

      I'll bet they'd also really appreciate some foreign (even dollar-backed) investment right now...

      --
      Downmodding is the refuge of the weak. Don't downmod, make a better argument!
    89. Re:Is this possible? by Anonymous Coward · · Score: 0

      The atmosphere is not a smooth surface; there are substantial convection currents well into the very high atmosphere from which extremely hot gas is ejected into space, forming the exospheric coma and tail. (The exobase is the point at which particle collisions approach zero, eliminating conduction as a heat transfer mechanism, and also the point at which particles follow ballistic trajectories, eliminating convective circulation (but not convection per se)).

      This planetary cooling process is so efficient that a small change to the albedo may lead to substantial real cooling of the whole planet.

      In short, we already have an excellent convective and conductive cooling system, rather than having to rely upon radiative cooling. Moreover, Earth's blackbody temperature is so low that it's not clear what current human-buildable materials could usefully dissipate it over a larger surface area.

      The atmosphere already does a great job.

  2. Environmental impact by phorm · · Score: 2, Interesting

    Uhhhh. Wouldn't making chips a bit more efficient be better, as opposed to making them "less likely to burn out at higher temps"

    Seems that google's not really thinking green in this case (despite the pretension to do so in others), unless they plan on making use of the datacenter heat elsewhere.

    1. Re:Environmental impact by LWATCDR · · Score: 4, Insightful

      Not really.
      No matter how cool the chips run they will put out heat. If you have two chips that run at X and use Y watts you will save power if you can run them a little hotter and use less power for cooling.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
    2. Re:Environmental impact by janeuner · · Score: 1

      So, you argue that they should use nothing but mobile chips because they use so much less power and generate a lot less heat?

      Never mind that you would have to build twice as many servers because of the performance penalty, or any other technical details for a distributed server. Just go with the "Google is bad" storyline - it works better.

    3. Re:Environmental impact by Rogerborg · · Score: 3, Interesting
      Ideally, yes. And ideally, I'd come home to find Alyson Hannigan oiled up and duct taped to my bed.

      Pragmatically, if they can't run cool, then it's more efficient to run them hot than to spend more energy actively cooling them.

      --
      If you were blocking sigs, you wouldn't have to read this.
    4. Re:Environmental impact by I.M.O.G. · · Score: 5, Insightful

      I'm not sure what your getting at - if by doing this they are saving millions in cooling expense, they are certainly using less energy. What is "going green" if it isn't energy conservation? The fact that the conservation comes from less work for their AC units rather than efficient little processors is immaterial.

      Don't expect any company to do things because its right - but good companies will find win-win situations where they cut costs and do things to "Go Green".

    5. Re:Environmental impact by Anonymous Coward · · Score: 0

      Both are good optimizations that you can do independently of the other. Efficiency is the obvious one and of course it's being worked on. I think the point here is that Intel has overlooked the second optimization up to now.

    6. Re:Environmental impact by gstoddart · · Score: 3, Insightful

      Seems that google's not really thinking green in this case (despite the pretension to do so in others), unless they plan on making use of the datacenter heat elsewhere.

      The amount of energy you need to use to cool that stuff is quite significant. And, in case you haven't realized it, generating cool air also create more warm air, it's just not in the data center. It's usually vented right outside.

      If the chips could run hotter, they'd have to use less energy to cool the data center, and generate less waste heat from that very cooling in the first place.

      I'm not convinced that what they're asking for isn't a good idea.

      Cheers

      --
      Lost at C:>. Found at C.
    7. Re:Environmental impact by SoupIsGoodFood_42 · · Score: 1

      Well, cooling them down takes energy and creates heat, so depending on how much you can save in that area vs. the extra expended with the chips, it might end up being more efficient overall.

    8. Re:Environmental impact by phorm · · Score: 1

      Uh, if the chips ran cooler, wouldn't there be less heat to dissipate in the first place?

    9. Re:Environmental impact by gstoddart · · Score: 2, Interesting

      Uh, if the chips ran cooler, wouldn't there be less heat to dissipate in the first place?

      Yes, but which is easier: making the chips more efficient, or allowing them to run a little hotter without melting?

      I honestly don't know. My first thought is that efficiency is harder than durability, but that's pulled completely out of my backside.

      I still think they're right in asserting that if they could handle a little more heat, then data centers would spend less energy trying to cool them to their operating range.

      Make them both more efficient (so they generate less heat) and run hotter (so they're less sensitive to that heat) and it seems like you win on two ends, no?

      Cheers

      --
      Lost at C:>. Found at C.
    10. Re:Environmental impact by Anonymous Coward · · Score: 0

      Even if they use the most efficient chip out there, the chips will still generate heat.

      The money is being saved by reducing expenses because is costs a lot of money to cool those chips down by an extra 5 degrees.

    11. Re:Environmental impact by bendodge · · Score: 0, Flamebait

      What is "going green" if it isn't energy conservation?

      Most peoples' idea of 'going green' somehow involves the government.

      --
      The government can't save you.
    12. Re:Environmental impact by moderatorrater · · Score: 5, Funny

      And ideally, I'd come home to find Alyson Hannigan oiled up and duct taped to my bed

      You know you're pathetic when they're even unwilling in your fantasies.

    13. Re:Environmental impact by phorm · · Score: 1

      To run chips cooler, you have to slow them (do not want) or to undervolt them.

      The "undervolt" option would fall in the efficiency category. The trick would be to make them perform at the same level (in terms of operations processed/time) while supplying less voltage. As it often seems that CPU makers tends towards "how can we crunch out more 'power' regardless of consumption before meltdown" as opposed to "how can we do the same amount of work in a similar amount of time with less power consumption," that would seem better.

      I don't see how lowering the power requirements of a chip reduces thermal tolerance, though.

    14. Re:Environmental impact by Surt · · Score: 1

      Uhhhh. Wouldn't making chips a bit more efficient be better, as opposed to making them "less likely to burn out at higher temps"

      Seems that google's not really thinking green in this case (despite the pretension to do so in others), unless they plan on making use of the datacenter heat elsewhere.

      Yes, assuming those were the tradeoffs, which they weren't.

      Google is trading existing performance levels against reduced cooling, which is a pure win, just by demanding more resilient chips from Intel.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    15. Re:Environmental impact by aperion · · Score: 2, Interesting

      It doesn't sounds like Google is asking for less efficient chips, that's a retarded notion. Instead it sounds like Google is asking for more durable chips.

      One that still operates at X% efficiency but operates at a higher ambient temperature., say 70F instead of 65F. I'm sure Google would like it better if the chips didn't produce any heat (ie 100% efficient), but that's impossible.

      Still, if you want to "Go Green" and be environmentally friendly, stop viewing heat as a problem. It's better to try and reclaim some of the lost energy (heat IS energy) than it is to spend more money trying to get rid of it. ie use energy to get rid of energy.

    16. Re:Environmental impact by Surt · · Score: 5, Insightful

      And ideally, I'd come home to find Alyson Hannigan oiled up and duct taped to my bed

      You know you're pathetic when they're even unwilling in your fantasies.

      You're assuming everyone prefers them willing.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    17. Re:Environmental impact by AmberBlackCat · · Score: 1

      The oil just makes it easier to free yourself from the duct tape.

    18. Re:Environmental impact by TheRaven64 · · Score: 1

      It's all about the temperature difference. The hotter something is, the more heat it will radiate and the faster it will cool. The bars on an electric heater get hot quickly and then stay the same temperature, because all of the energy being put in to them is being radiated away. A CPU is like this, only at a certain temperature it will fail. If you can bump up the temperature at which it fails a little bit then you can run it hotter, which means you need less cooling on the chip to stop it failing. More importantly, it means you can have the same amount of on-chip cooling but a higher ambient temperature. This means that your air conditioning needs to work less hard, and you might even be able to passively cool your entire data centre. The heat transfer from the internal air to the outside, if you pile the internal air through a radiator, is dependent on the temperature difference. If the internal temperature is five degrees more, then you can dump a lot more heat without having to go to active cooling.

      So, somewhat counter-intuitively, if chips ran cooler then you might actually need to spend more on cooling them.

      --
      I am TheRaven on Soylent News
    19. Re:Environmental impact by Anonymous Coward · · Score: 0

      Maybe he's already learned the hard way not to let her bring a musical instrument to bed, and the duct tape is also to keep her from telling stories about band camp.

    20. Re:Environmental impact by Asic+Eng · · Score: 1

      You seem to equate tied-up with unwilling - you're missing out on something really fun.

    21. Re:Environmental impact by Anonymous Coward · · Score: 0

      you're assuming she's unwilling; perhaps she duct taped herself, or had her friend do it for her!

    22. Re:Environmental impact by Anonymous Coward · · Score: 1, Insightful

      Not really.
      No matter how cool the chips run they will put out heat. If you have two chips that run at X and use Y watts you will save power if you can run them a little hotter and use less power for cooling.

      Electronics tend to use more power, the hotter they run. So running the chips at a higher temperature will use more power than at the stock temperature. However, since refrigeration based cooling can use up to 3 watts to remove 1 watt worth of heat, Google will still be coming out ahead by a long shot.

    23. Re:Environmental impact by jank1887 · · Score: 1

      no. it's duct tape. the oil does nothing.

    24. Re:Environmental impact by daft_one · · Score: 0

      She can't be too unwilling if she's still there, having been oiled up & then duct taped down. Think about it ;-)

    25. Re:Environmental impact by Firehed · · Score: 1

      Yes, but fixing the problem (inefficient silicon) rather than the symptoms (hot chips) not only has the same effect on cooling costs, but also lowers the initial power requirements. So not only do you not have to crank the AC as much to maintain $safe_operating_temperature, but you're sucking down less juice to power the chips at that temp.

      It's like replacing incandescent bulbs with CFLs, except that we want fast datacenters, not really bright ones (and without the nonsense whining about the tiny bit of mercury in the base of a CFL while forgetting that 10x as much is used in generating the power for a tungsten bulb). Sure, being able to run tungsten hotter without it melting on us would help the brightness issue, but it would still be sucking down more power and kicking off more heat. Invest in the R+D for the CFLs and you can get 5x the performance for the same power while not creating a furnace.

      Obviously the numbers don't translate between the two concepts, but the basic idea is about the same. Other than the fact that the R+D is already done on CFLs, while you're looking at billions to be spent on creating ultra-efficient processors. But for what it's worth, we already did it once at the consumer level from moving to the Pentium 4 architecture over to the P3-esque Core architecture.

      --
      How are sites slashdotted when nobody reads TFAs?
    26. Re:Environmental impact by Firehed · · Score: 2, Interesting

      I expect that increasing efficiency is considerably more R+D-intensive than just increasing tolerances (see: Pentium to Core architecture transition), but the latter may make a decent short-term solution until the former can be implemented.

      Of course, it's not just the processors that would need higher tolerances. Hard drives, while not generating nearly as much heat (or consuming as much energy) tend to be fairly picky, and as mechanical parts are probably much harder to improve tolerances, that could quickly become problematic even if Intel provided chips that could run pegged at 150c 24/365. I assume Google's servers like most are running ECC RAM so it may be less of a problem, but there are plenty of heat-sensitive components in computers besides the processors.

      --
      How are sites slashdotted when nobody reads TFAs?
    27. Re:Environmental impact by jonaskoelker · · Score: 1

      You're assuming everyone prefers them willing.

      You're assuming Alyson isn't willingly duct taped to the bed.

    28. Re:Environmental impact by Anonymous Coward · · Score: 0

      I just assumed he was really into duct tape.

      This is a site for nerds, afterall.

    29. Re:Environmental impact by Surt · · Score: 1

      Technically, I'm assuming she didn't lie to me about not being into that.

      --
      "Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
    30. Re:Environmental impact by AmberBlackCat · · Score: 1

      If I ever find myself trapped in a slashdotter's basement, I'll try to talk him into using some other type of tape instead. Perhaps VHS.

    31. Re:Environmental impact by LWATCDR · · Score: 1

      What are you talking about???
      What does CFLs have to do with this.
      Here let me put it in simple terms.
      1. Pick the cpu that has the highest efficiency.
      2. Get the manufacture to cherry pick CPUs for you that are have a 4C higher max temp.
      3. Use less power for cooling.
      4. Total power use is decreased.

      It is a lot faster and cheaper to get CPUs that are stable at higher temps than it is to create even more power efficient chips.
      Intel and AMD are already trying to make them as power efficient as possible. This is just a way to save a little more power and a little more money.

      --
      See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
  3. Underclocking if you're poor? by Junior+J.+Junior+III · · Score: 3, Interesting

    If you don't have the clout of a Google-sized organization to buy higher-rated chips from Intel, I wonder if you can basically achieve the same thing by underclocking. An underclocked chip will run cooler, but I don't know if it'll run more stably at higher temps, although I think it would.

    Does anyone have any experience with doing this?

    I think it'd be interesting to see whether the cost savings in power and cooling is offset by the cost of the performance losses.

    --
    You see? You see? Your stupid minds! Stupid! Stupid!
    1. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      Underclocking would probably cause the chip to generate less heat. So while it may not tolerate higher temperatures, the server farm would be easier to cool

    2. Re:Underclocking if you're poor? by LaminatorX · · Score: 3, Informative

      In the past, I've under-clocked primarily for noise benefits. Lower clock->lower temp->slower fam RPM->lower dbSPL.

    3. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      no, it would just mean a lower heat output.

    4. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      This is true. I remember when the fan on my K6-2 350mhz broke so I just ran it at 300mhz fanless.

      Easily more cool than running at full speed, and likely faster than just running a slower chip at full speed.

    5. Re:Underclocking if you're poor? by cduffy · · Score: 1

      Several years ago I underclocked my home system (which was initially quite unreliable) and saw a substantial decrease in uncommanded reboots. The issue went away almost entirely (even at full speed) when I upgraded the power supply; I suspect that I was running near the edge of what it could handle, and underclocking (by reducing the CPU's power draw) moved its usage far away enough from that boundary to make a difference.

    6. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      I can't believe my eyes! Someone on Slashdot is actually UNDERclocking a chip? You're going to have to give up your Slashdot membership.

    7. Re:Underclocking if you're poor? by I.M.O.G. · · Score: 1

      As an overclocker, your statement tends to be true in practice. The cooler a chip is kept, the closer you can get to its maximum overclocking frequency - the frequency beyond which the chip exhibits instability. Similarly, the lower the frequency is set the warmer temperature the chip will generally handle with stable operation. These are general trends, processors from different fabs or batches perform differently - but within the same batch of processors, you can reliably test and observe these results.

    8. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      My understanding is that Google has already done this for some generations of CPUs.

    9. Re:Underclocking if you're poor? by IorDMUX · · Score: 1

      When referring to CPU's like this, there are two different flavors of failure.

      First, there is the failure that causes your computer to crash--an improperly processed command, a bad memory jump, failure of the clock to get to the right places at once, etc. You reboot the computer and it will work again (for a bit, at least). Secondly is the physical breakdown failure that results in a broken CPU that will never work properly again.

      When you are looking at the risks/benefits from overclocking or underclocking, you need to be clear on the failure modes. Simply turning up the clock speed makes the first failure much more likely by reducing the leeway for a signal to get from point A to point B. So what do you do in this situation? Well, most users increase the CPU voltage to increase the driving strength of the signals. However, the heat generated on-chip is proportional to Vdd^2 as well as frequency^1. Operating in an increased heat environment, on the other hand, increases the odds of the second (catastrophic) failure [though it does slightly increase the odds of the first, as well].

      To get back to your original question, if you want to save on cooling and power with a CPU, you really need to under-volt it. Unfortunately, dropping the headroom *significantly* increases the chances of the first kind of failure as the default voltage has already been pushed surprisingly low by the manufacturer. (And besides, most consumer motherboards don't allow undervolting, anyways).

      --
      >> Standing on head makes smile of frown, but rest of face also upside down.
    10. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      Def not, if they clock to roughly twice the speed, there is no way it is going to get twice as hot. think about it

    11. Re:Underclocking if you're poor? by Anonymous Coward · · Score: 0

      Back when we had unwittingly purchased a Pentium 233Mhz I would underclock to 225Mhz by raising the bus speed and reducing the multiplier.

      The increase in bus improved performance more than having a higher overall clock speed. That also reduced the temps, but cutting a few degrees off wasn't as important as improving performance.

      Had I been smarter I'd have purchased the 200Mhz and overclocked it instead.

  4. Temp specs by MyLongNickName · · Score: 1

    Not mentioned in the story. What CPU are they talking about, and what is the upper end Google is looking for?

    (and this having to wait five minutes between posts is moronic. Look at my posting history, and all of them from the same IP address. Tell me why I have to wait this long to post.)

    --
    See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
  5. No one mentions a more obvious approach. by Anonymous Coward · · Score: 1, Interesting

    Under-clocking them a bit can't be that hard to do.

    1. Re:No one mentions a more obvious approach. by SoupIsGoodFood_42 · · Score: 3, Insightful

      That means you'd need to make up for the lack of processing power with additional CPUs, which would mean more CPUs to cool.

    2. Re:No one mentions a more obvious approach. by Z00L00K · · Score: 1

      Maybe it's time to think more of the performance/power consumption ratio when designing servers.

      More CPU:s may actually not be that bad because you can spread the dissipated power over a larger area. However you will also have larger computers.

      One way around it could be to locate datacenters at locations with natural cooling available like rivers and larger lakes.

      Today many cooling systems are aircooled, but the air can be a lot warmer and not able to absorb as much heat as water.

      --
      If builders built buildings the way programmers wrote programs, then the first woodpecker would destroy civilization.
    3. Re:No one mentions a more obvious approach. by Silentknyght · · Score: 2, Interesting

      Water is an excellent heat sink, but any company would likely run into serious environmental backlash if they wanted to use a lake or river as their heat sink. Just like on land, organisms in the water can be seriously disturbed by a change in temperature of even a few degrees. If the waste heat is seriously that large a problem, I'd recommend a man-made water cooling solution like a cooling tower, not too dissimilar from what goes on a power plants. Of course, most industrial or utility cooling tower sizes and appearances don't give off that special "nuclear" feel.

    4. Re:No one mentions a more obvious approach. by SoupIsGoodFood_42 · · Score: 1

      More CPU:s may actually not be that bad because you can spread the dissipated power over a larger area.

      But that doesn't equate to less heat dissipated in total. While I'm sure such things influence the design of a cooling system, without any specifics, there is no reason to assume it's an advantage.

    5. Re:No one mentions a more obvious approach. by hierophanta · · Score: 1

      Under-clocking them a bit can't be that hard to do.

      Under-volting them a bit cant be that hard to do.
      there fixed that for you (always wanted to say that)

    6. Re:No one mentions a more obvious approach. by tyrione · · Score: 1

      When Nuclear Power generation finally switches to Pebble bed Modular Reactors [first invented in the US and blocked by the Atomic Energy Commission (1944) for it's lack of military options] using Liquid Cooled Helium then I'm sure someone would rather adapt a small-scale solution, using a material medium other than water, for motherboards to do this than just using a giant water sink to account for the data center, as a whole.

      ME Magazine: Pebbles making Waves, ME Magazine

      PBMR Corporation

      US Westinghouse takes stake in PBMR

    7. Re:No one mentions a more obvious approach. by Henneshoe · · Score: 3, Funny

      What they should do is get the processors to run at 100 degrees C. Then they could use the heatsinks to boil water, the steam to turn a turbine and the turbine to generate electricity. It would be somewhat like the turbocharger for your car.

      Slashdot Post Checklist
      Bad Car Analogy - Check
      Spelling Errors - Check
      A solution that won't work in real world - check
      Reply to a comment on top 10% of page so people read yours - check

    8. Re:No one mentions a more obvious approach. by camperslo · · Score: 1

      But that doesn't equate to less heat dissipated in total. While I'm sure such things influence the design of a cooling system, without any specifics, there is no reason to assume it's an advantage.

      If it is exceeding a specified peak temperature at the chip, spreading the same amount of energy out certainly would help. Improvements in the package design for the CPU (lower thermal resistance) might be able to lower the chip temperature sufficiently without changing the amount of power radiated. Of course improving the heat flow from the chip package to heatsink, or heatsink to ambient air, would have a similar advantage. Those methods would not require an increase in the number of systems or CPUs.

      Whether Intel could actually have the chip run hotter depends of what the failure modes are. Elevated temperature generally increases the long term failure rate even if it doesn't cause short term failure. If Intel could get away with actually reducing core voltage (and thus power) slightly while still supporting the specified clock, that would be preferable to running hot.

      If a design pushes CPUs closer to thermal maximums to reduce air conditioning costs, it is important to note that the time-to-failure in the event of a cooling system problem is reduced. One should plan for backup cooling, down-clocking or some other means of functioning during a cooling system malfunction.

    9. Re:No one mentions a more obvious approach. by spectecjr · · Score: 2, Informative

      Water is an excellent heat sink, but any company would likely run into serious environmental backlash if they wanted to use a lake or river as their heat sink.

      Oh, I dunno, Bethlehem Steel in Indiana (just off of Lake Michigan) did that for years. It was nice going through the outflow in a Hoby Cat... drag your feet through the warm water and enjoy. Good fishing there too.

      --
      Coming soon - pyrogyra
    10. Re:No one mentions a more obvious approach. by flibbajobber · · Score: 1

      How is this fixed? A modern CMOS chip "consumes" energy during the clock transition - when each FET's gate capacitance needs to be charged or discharged. Once this occurs there is virtually zero current drawn by the chip.

      So power draw is almost entirely proportional to clock frequency.

      To be fair though, I guess a certain operation requires a certain number of clock cycles, regardless of the speed at which they occur. So perhaps a slower rate lowers the *power* draw, but not total energy consumption (all else being equal)

      Best of both worlds (since all else never is equal) would be underclocking, thereby allowing a greater amount of undervolting (since gate charge would be proportional to supply voltage, that has a inversely proportional affect on energy consumption)

    11. Re:No one mentions a more obvious approach. by rtb61 · · Score: 1

      Then you just employ lateral thinking. Google can place it's data centres on the coasts in regions where there is a scarcity of fresh water. Your typical reverse osmosis desalination plant is only around 20% effective and the waste water is returned to the sea. Google just needs to implement a high volume heat exchange in the waste water stream and they can sell the fresh water. So heat loss at a profit rather than at a loss and, it is major cities in a lot of locations that are looking for fresh water, so customers, staff and network infrastructure are all readily accessible. So a bit of clear thinking and you don't have a ship adrift at sea but a for more financially viable solution that supports the local community.

      Now the amount of cooling water available will likely far exceed Googles requirements so that could sell access to it at a data park, where other companies with similar issues with only relatively low heat loss requirements could also gain access to low cost cooling solutions. Hmm, google, too many marketing and psychology PHDs and not enough engineering common sence.

      --
      Chaos - everything, everywhere, everywhen
    12. Re:No one mentions a more obvious approach. by hierophanta · · Score: 1

      this is how it is fixed: watts = volts * amperes
      its that simple. less voltage = less power consumption

    13. Re:No one mentions a more obvious approach. by aliquis · · Score: 1

      As long as you have enough air to cool the vapor why wouldn't it work? Just because Googles has issues with it not everyone does.

  6. Waste Heat reclamation by LaminatorX · · Score: 4, Funny

    When in college, I heated my crappy little schack by putting 150W bulbs in every light. It was like my own little Easy-Bake oven.

    1. Re:Waste Heat reclamation by Anonymous Coward · · Score: 0

      Funny, I just ran Folding@home on a bunch of Pentium IIIs.

    2. Re:Waste Heat reclamation by Abreu · · Score: 1, Troll

      I gather you were not paying the electricity bill...

      --
      No sig for the moment.
    3. Re:Waste Heat reclamation by entgod · · Score: 1

      Probably not, but even if he were, at least he wouldn't have been paying the heating bill. And he got extra light, of course.

    4. Re:Waste Heat reclamation by LaminatorX · · Score: 2, Informative

      The place was all electric. An incandescent bulb compares favorably to many space heaters in terms of V->heat efficiency, and you get light as a bonus.

    5. Re:Waste Heat reclamation by dsanfte · · Score: 1

      Uh, all electric heaters are 100% efficient. Take thermodynamics sometime.

      --
      occultae nullus est respectus musicae - originally a Greek proverb
    6. Re:Waste Heat reclamation by KevinIsOwn · · Score: 1

      Right, and that orange glow that space heaters tend to give off is simply magic left over from the thermodynamics pixies.

    7. Re:Waste Heat reclamation by LaminatorX · · Score: 1

      And yet, they sometimes glow, or have higher inductive losses due to large coils, or use part of their power for a blower, or...

      Learn the difference between theory and practice sometime.

    8. Re:Waste Heat reclamation by NerveGas · · Score: 1

      In one place I had during college, I didn't pay the electric bill... but the owner refused to allow decent heating. So I just left every electrical appliance on all of the time. It worked pretty well. :D

      --
      Oh, you're not stuck, you're just unable to let go of the onion rings.
    9. Re:Waste Heat reclamation by bloobloo · · Score: 1

      And when the photons hit a surface, the energy is absorbed. This is by conversion from light to heat. Therefore if you want heat, the light is just a transmission medium.

    10. Re:Waste Heat reclamation by Anonymous Coward · · Score: 0

      More like a terrarium! I did this after college... I rented a house with a few of my friends. The house has such high ceilings, was drafty, and a $300 tank of oil ran out in 3 weeks! So, my friend had the brilliant idea to install heat lamps in the light sockets in the ceiling. When the lamp struck your skin, it created warmth. It was like we were lizards in a terrarium. That and electric heaters got us through the cold PA/NJ Winter.

      Let me tell you, it was STILL cold as hell but we only had to fill up the oil tank twice! I remember one day were were playing that new Mortal Kombat game on XBOX 1, so we just sat in front of the electric heaters, basking in heat lamp light, and drinking a diverse set of liquor to stay warm! AHh memories...

    11. Re:Waste Heat reclamation by JanneM · · Score: 1

      The lightbulbs are mostly near the ceiling, so the heat (the hot air created, actually) mostly just gathers there and leaks out instead of being generated near the floor where it would actually do you some good before rising.

      Efficient, sure, but how's that helping when the heat is not where you ant it to be?

      --
      Trust the Computer. The Computer is your friend.
    12. Re:Waste Heat reclamation by dsanfte · · Score: 1

      A heater sucks a net wattage off the line.

      "And yet, they sometimes glow"

      Light turns into heat when it hits and is absorbed by a surface.

      "or have higher inductive losses"

      Just inhibits the circuit and lowers the heating wattage, has absolutely no effect on efficiency.

      "or use part of their power for a blower"

      This is also turned into heat (eventually).

      You see, I know the difference between theory and practice. There's absolutely NOTHING you can do with energy that won't eventually turn into heat, and electric heaters are always 100% efficient at what they do. Always.

      --
      occultae nullus est respectus musicae - originally a Greek proverb
    13. Re:Waste Heat reclamation by Anonymous Coward · · Score: 0

      No, still it will all go to heat. Any energy "loss" is a conversion to heat.

  7. Intel says they don't by the_skywise · · Score: 1

    If the chips failed prematurely at these higher temperatures, the former Googler says, Intel was obliged to replace them at no extra charge.

    Intel denies this was ever the case. "This is NOT true," a company spokesman said in an email. Google declined to comment on its relationship with Intel. "Google invests heavily in technical facilities and has dozens of facilities around the world with many computers," reads a statement from the company. "However, we don't disclose details about our infrastructure or supplier relationships."

    So Google claims they're more environmentally friendly... but burn through chips faster.

    1. Re:Intel says they don't by Anonymous Coward · · Score: 0

      Where does it say they burn through chips faster? All it says is that Intel replaced chips for free, there is no mention of a higher failure rate.

    2. Re:Intel says they don't by MozeeToby · · Score: 1

      You seem to think that 'being green' is a binary position; that every action is either good for the environment or bad for it and there is no ambiguity or overlap.

      Google runs their chips at higher temperatures and yes, that causes more than the average number of failures. Of course, Google data centers require much, much less cooling than their compition's data centers do. I'd be willing to bet that they save more power by reducing cooling costs, than they waste by requiring more CPUs to be manufactured.

      It's the same thing with all green technology, there are always trade offs. A new hybrid with a lithium ion battery is much more energy intesive to design and build than a car with a standard drivetrain. I'm not saying that hybrids are bad for the environment, I'm just saying that whether a technology is a net gain or loss for the environment isn't so easy to calculate as just looking at a single number.

    3. Re:Intel says they don't by the_skywise · · Score: 1

      I agree it's not a binary position.

      But depending on their burn rate, they may be at a wash for the energy they're saving compared to the energy spent to make new chips.

  8. Not too surprising by Ngarrang · · Score: 1

    When you are a big company that spends enough money, you can ask for this sort of thing and your demand will be met.

    "Guarantee us a higher temp CPU or we will switch to AMD...and tell everyone about it."

    I have a feeling that the CPUs can handle a bit more temp than they are rated as a CYA move by Intel, anywho.

    --
    Bearded Dragon
    1. Re:Not too surprising by stilwebm · · Score: 5, Insightful

      "Guarantee us a higher temp CPU or we will switch to AMD...and tell everyone about it."

      That's not really how the negotiation goes in this type of situation where there are two major supplier choices (AMD and Intel) and Google is a relatively small customer when compared with Dell, HP, IBM, etc.

      In all likelihood, the negotiation is more of a partnership where both parties work together to create value. Google says, "We buy thousands upon thousands of your chips, but we also pay millions of dollars annually to cool them. We'd be willing to pay little premium and lock in more volume if you can help us save money by increasing the temperature inside our data centers." Google has done the math and comes prepared with knowledge of how much those specs can save them and forecast of chip need of the next 12-18 months, and the two work together to create value. For example, Google might offer to allow endorsements (as they did with Dell for their Google appliances) in exchange for favorable pricing and specifications.

      The "do this or I'll switch" tactic only works well when there are many suppliers and products are relatively undifferentiated, like SDRAM or fan motors.

    2. Re:Not too surprising by fuzzyfuzzyfungus · · Score: 2, Interesting

      It also wouldn't surprise me if Google were willing to offer something of a testbed setup. A while back, they put out that report on HDD reliability and its influences, so they are obviously watching that sort of thing. And, since their historical style has been very much about redundant masses of commodity gear, they can theoretically tolerate slightly higher failure rates if those lower costs in other ways.

      I suspect that, with negotiation to set the correct balance of pricing, warranty, access to handpicked chips, etc. both Intel and Google could easily benefit from an arrangement where Google gets to play with slightly experimental stuff, like higher temperature processors, and Intel gets field reliability data.

    3. Re:Not too surprising by freedom_india · · Score: 1

      In reality that IS true.
      The Very Large bank where i was employed earlier had a special agreement with Microsoft.
      They got a customized version of XP meant specially for the bank's hardened network. Yeah i know the Admin kit allows customization, but i mean at a lower level: the NSA hookups in the system DLLs were not present!
      As soon as a Dell entered the bank, it was wiped out, and this OS was installed. It was a weird mix of little Linux bootup code and XP.
      You had all rights of an admin EXCEPT when it comes to modifying system32 or adding hookups to startup.
      Guess they had a different code base, because my laptop came with the bank's logo'ed recovery disc which had its bootup code and OS made from XP.

      --
      "Doing what i can, with what i have." ~ Burt Gummer
    4. Re:Not too surprising by npfscayle · · Score: 0

      i call bullshit, what NSA hookups?

    5. Re:Not too surprising by stilwebm · · Score: 1

      Good point, I hadn't thought of the high failure rate tolerance as a way to contribute to Intel's R&D. That's exactly the type of partnership value they'd seek.

    6. Re:Not too surprising by Mesa+MIke · · Score: 1

      The secret ones that the NSA doesn't want you to know about, of course.

  9. WTF? Lawyers as engineers, not so much by Ancient_Hacker · · Score: 3, Informative

    This sounds like a scenario where lawyers are trying to act as engineers. That works about as well as you might expect.

    There are these engineering things, amusingly called "Schmoo plots", that map out a chip's operating envelope of voltage versus speed versus temperature. From those an engineer can forsee how hot you can run a chip before its rise and fall time margins start to get marginal.

    There is very little Intel can do to stretch thing by another 5 degrees. It's not something that can be imposed by fiat. Intel engineers have already juggled all the variables to come up with the best performance possible. SOMETHING is going to have to give. Either the chips will have to be selected and graded for speed, lowering the overall envelope for the chips everyone else gets, or they'll have to fudge some other parameters, hoping nobody will notice, or worse yet they'll tweak some variable right to the edge of raggedness, resulting in worse reliability down the road.

    Lawyers and accountants generally don't know you can't have everything. let's hope the engineers educate them.

  10. I for one am not surprised. by nimbius · · Score: 1, Flamebait

    other businesses have this same questionable practice. for example, walmart requires special packaging from its suppliers that is not normally afforded to other retailers. broadcomm, microsoft, and nvidia likely have a few cozy agreements that are exclusive and hushed. a possible example here might be the ACPI standard and how it seems to "just work" in windows but struggle in some cases with *nix.

    it certainly gives google a cost advantage, and i can imagine why they vehemently deny it in TFA as i glance over the justice department article. although whatever gains google makes up for in cooling, they may just as easily have lost in a more power-hungry architecture overall:

    http://people.freebsd.org/~brooks/papers/bsdcon2003/fbsdcluster.pdf has experienced it, and his 2007 update also confirms.

    im left wondering what AMD might do for its biggest customers?

    --
    Good people go to bed earlier.
    1. Re:I for one am not surprised. by Anonymous Coward · · Score: 0

      im left wondering what AMD might do for its biggest customers?

      Knowing their customer base, AMD likely sends them a baseball cap and an overclocking guide.

    2. Re:I for one am not surprised. by rrohbeck · · Score: 1

      Dell did (does?) the same thing by having higher temperature specs for their servers than the rest of the industry. Of course customers will see higher failure rates if they actually use the larger margin.
      It's teh physics, stupid!

    3. Re:I for one am not surprised. by Anonymous Coward · · Score: 0

      The ACPI standard "just works in Windows" because BIOS manufacturers determine which parts of the ACPI standard that Windows cares about and then only write/test their code against that.

      Then when linux came around and wrote something that takes advantage of the entire ACPI standard, bugs in the underlying BIOSes are exposed.

      Blame the BIOS makers, not some vast MSFT conspiracy.

    4. Re:I for one am not surprised. by nimbius · · Score: 2, Informative

      According to microsoft antitrust evidence, it certainly seems like the conspiracy holds some water ;)

      http://antitrust.slated.org/www.iowaconsumercase.org/011607/3000/PX03020.pdf

      --
      Good people go to bed earlier.
  11. Why don't they by iminplaya · · Score: 1

    just buy Intel?

    --
    What?
    1. Re:Why don't they by TheRaven64 · · Score: 1

      Because Intel's market capitalisation is $89bn, while Google's is only $109bn. They don't have anything like the liquid assets required to purchase Intel.

      --
      I am TheRaven on Soylent News
    2. Re:Why don't they by iminplaya · · Score: 1

      Well, can't the government just give them another 700 bil? All they have to do is to threaten bankruptcy and that thousands of jobs will be lost if they don't.

      Just making sure that nobody misunderstands the nature of this post

      --
      What?
  12. Re:WTF? Lawyers as engineers, not so much by mindstrm · · Score: 5, Insightful

    Odds are this is being driven by a data-center engineering team, who are looking at the cost savings of running their data center 5 degrees hotter.

    You don't get what you don't ask for.

    Intel will do exactly as much engineering as necessary to keep their target market up, and no more.

    If the market wants chips that operate 5 degrees hotter.. the engineers will do their job and see if it can be done. Intel will charge a premium for this.

    That's business.
     

  13. Google Runs Its Data Centers at 80 Degrees by miller60 · · Score: 1

    Google said recently that it runs its data centers at 80 degrees as an energy-saving strategy, so chips that support higher temperatures would mean fewer hardware failures in their data center. Most data centers operate in a temperature range between 68 and 72 degrees, and I've been in some that are as cold as 55 degrees. Lots of companies are rethinking this. In the Intel study on air-side economizers, they cooled the data center with air as warm as 90 degrees. ASHRAE is also considering a broader temperature range for data center equipment.

    1. Re:Google Runs Its Data Centers at 80 Degrees by Skapare · · Score: 1

      You can run a data center cheaper at a cooler temperature simply by having better insulation. The rate of energy in (power) equals the rate of energy out (power) at any stable temperature. But there is more energy in than just the electricity being used. Sunlight on the building is also energy in, to the extent that it bypasses the insulation. Better insulation limits this energy going in. If the insulation reduces the energy in from outside at a rate greater than the energy used in the form of electricity needed to pump heat around (blowers, etc), then you have a win and the stable temperature goes down.

      Locating a data center on the north side of a mountain and/or at latitudes closer to the poles, is another way to reduce solar input leakage. But at the same time, it's also less opportunity for free solar electricity. One location, Iceland, does offer the cooler northern climate, reduced solar impact, and cheap geothermal electricity. They just need to build some big pipes (to North America and Europe). It could be a great place to put in an "extra capacity on demand" data center that could be used for European lookups when it peaks there, then North American lookups when the peak shifts over a few hours later, and then do crawls the rest of the time.

      --
      now we need to go OSS in diesel cars
    2. Re:Google Runs Its Data Centers at 80 Degrees by mabhatter654 · · Score: 1

      you have to remember that if the room is 80 degrees the processors are probably at 110 degrees while exchanging heat. That's the number Google wants raised. There's big bucks in cooling data centers. I probably costs several times as much to power and cool a chip over it's life than is paid for it. Expecting Intel to put extra engineering will save lots of electricity down the road.

    3. Re:Google Runs Its Data Centers at 80 Degrees by the+eric+conspiracy · · Score: 2, Insightful

      You can run a data center cheaper at a cooler temperature simply by having better insulation.

      That assumes that the outdoor temperature is higher than the indoor temperature. My bet is that a data center run at 80 F in the Pacific Northwest would be warmer inside for most of the year than outside. Insulation under those conditions could actually increase cooling costs.

  14. Commercial or industrial scale? by ajb44 · · Score: 1

    Hmm, they don't say if this is commercial (0..70) or industrial (-40..85) temperature range - I guess intel chips are normally commercial range, so they've bumped then up to 75.

  15. Why doesn't Google just buy Intel? by bugeaterr · · Score: 1

    Then they can set the temperature to whatever they want. ;)

  16. Are they saving MILLIONS? by hackus · · Score: 5, Interesting

    Most of the power supply systems for my servers, which are HP G3-5 systems of various U sizes, tend to waste more power as temperature goes up.

    This has nothing to do with CPU's though. It is the power supplies on the machines. As temperature goes up, efficiency goes down. At around 80 degrees I noticed a significant larger draw on the power supply with my amp meter.

    I had a gaming system with two ATI 4870's and the 800 Watt power supply would crash my machine if I did not run the air conditioner and keep the room at 70 degrees after some fairly long Supreme Commander runs.

    I noticed that the amperage would go up, and the power output would go down as temperature would go up.

    I have not conducted any experiments in a lab setting with this stuff, but from experience, jacking the temperature up usually makes power supplies work harder and makes them less efficient.

    -gc

    --
    Got Geometrodynamics? Awe, too hard to figure out? Too bad.
    1. Re:Are they saving MILLIONS? by Penguinoflight · · Score: 2, Informative

      Traditional 120-many voltage DC power supplies certaily suffer from lesser efficiency at higher temperatures. Running two 4870s on a single 800w power supply probably isn't a good idea, especially if you have a high-powered CPU to go with them. Most quality power supplies will be rated lower than their maximum output to allow for temperature concerns.

      These things said, google uses custom power supplies and systems that only run on 12v. These power supplies may be easier to generate in quality batches, but will still be subject to the same efficiency curves.

      --
      "And we have seen and do testify that the Father sent the Son to be the Savior of the World"
      1 John 4:14
    2. Re:Are they saving MILLIONS? by ChrisA90278 · · Score: 1

      Do ypu think maybe that Google is using DC power? That way they have just a few large power supplies and another room and send DC power over large copper bus bars to the racks. These DC systems are expensive but you make the money back in power/cooling and you save money with the UPS too.

    3. Re:Are they saving MILLIONS? by wvmarle · · Score: 1

      Somehow I doubt datacentres like the ones Google operates use switching power supplies, located next to the hardware they power, like in your home computer. I for one would consider building a single power supply pushing a lot of amps through some fat cables that branch off to where-ever power is needed.

      But then I've never seen a datacentre from the inside, so I may be totally wrong.

    4. Re:Are they saving MILLIONS? by Toll_Free · · Score: 1

      As your power supply is a switching type supply, and the demands at low power cause the power supply itself to actually sink MORE current (causing the thing to get warmer), I think your overheating problems where more towards the machine and heat, rather than power supplies and heat.

      Switching type supplies are nothing more than switches. They turn on and off at incredible speeds.

      This makes them way more efficient than linear, which are always on. Switchers have the benefit of turning on and off and during the off time (when the power supply is actually supplying current to the load), the transistor(s) are sitting still, resting. During the on time, they are conducting and sinking current to ground, causing them to run hotter.

      Your logic is right on. Your physics is right off.

      The power supply will run warmer when sinking more current. The wiring will run warmer when drawing more current. Nothing is free.

      Having to have a computer that requires you to maintain your living room / bedroom / computer room at a certain temperature is crazy.

      --Toll_Free

    5. Re:Are they saving MILLIONS? by Anonymous Coward · · Score: 0

      As temperature goes up, efficiency goes down. At around 80 degrees I noticed a significant larger draw on the power supply with my amp meter.

      That's all true, but there are more energy conversions involved in cooling than there are in the hardware (electricity into heat -> increased resistance.)
      It is costlier in energy and money to cool the whole setup down 5 Kelvin then to take on a rise in amperage because of +5 Kelvin heating causing higher resistance.

      Given the scale of their operations, this saving them millions sounds plausible.

  17. But they pass it off to someone else by Moraelin · · Score: 5, Insightful

    Yes, but way I see this is:

    Intel isn't arbitrarily going, "man, we could make chips that run ok 5 degrees hotter, but we're gonna piss everyone off by demanding more cooling. Just because we can." Most likely Intel is already doing the best it can, and getting a bunch of chips which vary in how good they are. And they're getting the same bunch of chips regardless of whether Google demands higher temps or not.

    Google just gets a cherry-picked bunch, but the average over Intel's production is still the same. Hence everyone else is getting a worse selection. They what remains after Google took the best.

    It's a zero-sum game. The total load on the planet is the same. The same total bunch of chips exits Intel's fabs. On the total, no energy was conserved.

    So Google's "going green" is at the cost of making everyone else less "green". They can willy-wave about how energy efficient they are, by simply dumping the difference on someone else.

    That's not "going green", that's a predatory approach. Your computers could require on the average an extra 0.5W in cooling, so Google can willy-wave that theirs uses 1W less. They just dumped their costs and their "eco burden" to someone else.

    It's akin to me willy-waving that I'm so green and produce less garbage than you... by dumping some of my garbage in random other people's garbage bins across the town. Yay, I'm so green now, you all can start worshipping me. Well, no, on the total the same amount of garbage being produced, I just dumped it and the related costs on other people. That's not going green, that's being a predator.

    I can see why a business might want to cut their own costs, and not care about yours. That's, after all, the whole "invisible hand" theory. But let me repeat it: on the whole no energy was conserved. They just passed off some of their cooling costs (energy _and_ money) to someone else.

    --
    A polar bear is a cartesian bear after a coordinate transform.
    1. Re:But they pass it off to someone else by Muad'Dave · · Score: 5, Insightful

      I agree with most of what you said, but I think there is a _slight_ difference between Google having the higher-temp-rated chips and your average Joe user having them. Google's chips will be running full throttle/full temp 24/7; Joe user might run them full blast (and therefore full temp) 2% of the time. I bet the energy savings are not insignificant when usage patterns are taken into consideration.

      --
      Tiller's Rule: Never use a word in written form that you've only heard and never read. You will end up looking foolish.
    2. Re:But they pass it off to someone else by aniefer · · Score: 1

      It's a zero-sum game. The total load on the planet is the same. The same total bunch of chips exits Intel's fabs. On the total, no energy was conserved.

      So Google's "going green" is at the cost of making everyone else less "green".

      Not true.

      This is only the case if you assume every other customer is also running data centers that require special cooling. If Google is using off-the-shelf components, then it is just as likely that the other chips are going to regular desktops that receive no additional cooling.

      In this case, it is actually more efficient overall to give the best chips to Google and other datacenters, and leave the rest to others.

    3. Re:But they pass it off to someone else by mabhatter654 · · Score: 2, Insightful

      you miss the point entirely. Google wants Intel's processors to operate properly at a higher temperature than currently spec'd that will allow Google to use less cooling. They want the processors to tolerate more heat, not generate more heat. This means Intel will have to make the processors slightly more beefy, but how much does that really cost across millions of units once the design work is done, a few bucks per processor tops.

      Google pays dozens of dollars a MONTH to cool each processor. Intel making this change may cost them a dozen dollars up front, one time. If Intel spends a little extra up front to make a processor that allows higher temps, Google will save multiple times the Processor's cost in electric bills... that's REAL efficiency at work. Google wants Intel to design a chip $20 more expensive so they can save $1000's in energy cost.

    4. Re:But they pass it off to someone else by Anonymous Coward · · Score: 0

      No Google is not getting a cherry picked bunch. All Google is getting is a warranty that if the chip fails while using at 0-5c higher then the normal published spec, Intel will cover it.

      More than likely, the CPU temp spec is the limiting factor in their "computers" and the only thing preventing them from changing their room temp. They don't want to do it and not have coverage so they want something from Intel that says go ahead.

      For everyone making the claim of cherry picking, can you provide ANY basis on how or what those chips are singled out and is this something Intel does as a normal course of business? I've heard they put some through different clock tests so maybe they actually do? Again, some proof, not speculation or assume that it happens because everyone else says it does so it must be true.

    5. Re:But they pass it off to someone else by beckje01 · · Score: 1

      I would have to agree with you, usage patterns will play into this a lot. But lets look at this another way.

      Google takes the best chips, those chips are being run hotter then Intel recommends for the rest of the batch. So if Google wouldn't do this the cooling required would be related to Intel's recommendation by Google demanding a subset of chips that can run hotter they create a set of chips that require less cooling then Intel's recommendation. Thus the net cooling required is lower then if just using Intel's recommendation.

      In the purest sense the GP makes sense but in the real world where the cooling cost is based on Intel's specs not what the chip can do Google is lowering the overall cooling cost, assuming 100% load at all time. But with usage patterns they are lowering the cost on a big segment that will run at 100% load while most other chips will not be running at 100%.

    6. Re:But they pass it off to someone else by Anonymous Coward · · Score: 0

      Sorry, you are wrong. It's not a zero sum game, Google is actually changing things.

      Intel produces a batch of chips. If they don't change the spec or do anything special, everyone has to design their computers and data centers around the chip spec. Some chips can run hotter than spec, but since there is no sort it doesn't matter, all chips must be cooled to the same spec.

      Now Google comes and buys a small % of the chips from this batch. Intel can easily sort the chips that can do this so they pull them out and ship to Google. Google runs them 5 degrees hotter and saves money.

      The remainder of the batch get sent to the rest of the customers who run them at the initial spec.

      Net - Google's chips run hotter, they save money on cooling, the planet is better. Nothing else is changed.

      Remember - a chip temperature spec says that EVERY SINGLE CHIP you buy will work at xx degrees. Intel then guardbands that by a few degrees to make sure they meet the spec (that's why you can overclock). When they test temperature range they can sort for the extra 5 degrees and send them to Google.

      Processor speed is sorted the same way, but since it is a parameter that makes a difference to the end product Intel charges (sometimes REALLY CHARGES) extra for the faster chips. They all come from the same design and wafer, they just happen to run faster.

    7. Re:But they pass it off to someone else by TheLink · · Score: 1

      > So Google's "going green" is at the cost of making everyone else less "green".

      That's not true if the chips just tolerate higher temps better (instead of generating less heat - which is a different thing).

      Giving chips that can't take higher temps to Joe Average is not going to increase energy consumption - since:
      1) Joe Average is going to have air conditioning/heating set for the humans
      2) Joe Average's house/room does not have a high server density and has poor insulation, so most of the airconditioning energy is used to pump out heat that leaks in from outside, rather that heat from the few computers on in Joe's house/room.

      Conventional Wisdom Company is going to have their server air conditioning set to "way cold", either out of habit or they really have stuff that really needs low temps (there used to be stuff that caught fire rather easily :) ).

      So no difference to them.

      That said the CPUs can take about 60-70C, so it's not a big deal even if ambient is 40C (104F), as long as the airflow is enough to take the heat away.

      So I wonder what temperatures the chips are running at in Google's data centers - some articles claim they have ambient at 80F (26C), what are the temps at motherboard and CPU? If the density is high enough, those temps might be high...

      I'm in a tropical country and the computers in my home are definitely exposed to an ambient of > 80F. Heck 86F-92F (30C-33C) wouldn't be far from typical. And they work fine.

      I think countries in the Middle East region are hotter - doubt all PC owners there have airconditioning.

      I'm actually more worried about the hard drives failing due to high temps.

      --
    8. Re:But they pass it off to someone else by poot_rootbeer · · Score: 1

      They want the processors to tolerate more heat, not generate more heat.

      And where does the heat that the CPU needs to tolerate come from? From inside the CPU itself!

      We're not talking about a reverse-heatsink scenario here where the ambient temperature of the air inside the case is hotter than the CPU core itself.

    9. Re:But they pass it off to someone else by Anonymous Coward · · Score: 0

      Not to mention, not everyone has closely regulated temperatures in their homes... if my machine runs a little hotter, my A/C isn't going to work any harder (I don't have A/C) - most likely, the difference wouldn't be noticeable, but if it were I may just open a window.

      I'd assume the greatest benefit for lower-temp processors (even granted that means hotter chips for the rest of us) would be had where cooling systems are professionally installed and micromanaged, where systems are run at peak load 24/7, and where economies of scale may kick in.

      In short, it's NOT a zero-sum game.

    10. Re:But they pass it off to someone else by steelmaverick · · Score: 1

      Well, you have to put them into context. The average user, with one CPU, isn't going to see much of a power usage savings advantage because 5 degrees to a single user on a single computer is minuscule. Put that to google's data centers, where they run tens of thousands of servers in a single room at 100% usage 24/7, and the power savings will be HUGE because they're being far more efficient. Besides, most processor coolers try to make the CPU run cooler anyway.

      --
      Proudly posting without RTFA.
    11. Re:But they pass it off to someone else by nabsltd · · Score: 3, Interesting

      Google's chips will be running full throttle/full temp 24/7

      Is there any documentation for this?

      I seriously find it hard to believe that Google has every processor they own running 24/7 at 100% utilization. Other than the computation problems like SETI and protein folding, most problems are I/O bound, and I would think that the stuff Google does would involve a lot of I/O.

    12. Re:But they pass it off to someone else by nabsltd · · Score: 1

      And where does the heat that the CPU needs to tolerate come from? From inside the CPU itself!

      You are still missing the point.

      Any CPU running even 25% load will quickly build up a lot of heat inside the CPU package if the heat isn't moved away. This is what heatsinks and fans are for.

      Google wants a CPU package that can tolerate having less of the heat moved away (or moved less slowly) into the outside air. This means that the air doesn't have to be as cool to keep the processor within an acceptable temperature range. This, in turn, translates into less air conditioning used, which means less power used.

      The CPU uses the same amount of power as before, so overall it's a power reduction.

    13. Re:But they pass it off to someone else by Asic+Eng · · Score: 1
      Not really - they just want to select a set of chips which can stand higher ambient temperature. Using those they can set their AC to 5 degrees higher. This only works if every single chip in the room can withstand the higher temperature. If you have two data centers, and each of these has 50% chips which can run at higher temperature, and 50% which can't - then you need to set the AC in both places low enough for the 50% of chips which need it.

      So the savings in one data center are only possible if the chips are concentrated there - the second data center doesn't spend anything extra.

      I'm a chip designer, one of the things we do for our chips is burn-in testing: we test that they can run at high temperature for a long period of time. It's conceivable (though by no means certain) that most of Intel's chips can tolerate the higher ambient temperature, they just aren't tested for it. In this case Intel would merely add another burn-in step (or modify the existing burn-in).

      Some chips dissipate a lot of heat, some chips have strict requirements for ambient temperatures. These are different issues - they are related, but you can't in general say that a chip dissipating more heat also has stricter requirement for ambient temp.

    14. Re:But they pass it off to someone else by Anonymous Coward · · Score: 0

      How is a chip that is able to operate 5 degrees hotter requires more cooling?
      The chips are the same, they generate the same amount of heat. What google wanted is certification that its ok to cool them less so they operate 5 degress higher with no problems. Seem a pretty good idea to me.
      its like having a cpu and the specs say that it is safe to work up to 85 degrees and you ask "give me the same but make it safe till 90 degrees so i can use the less expensive cooler and still have warranty"

    15. Re:But they pass it off to someone else by jank1887 · · Score: 1

      I'll email google and ask for some of their company proprietary server operating statistics for use in a public slashdot discussion.

    16. Re:But they pass it off to someone else by hierophanta · · Score: 1

      Yes, but way I see this is:

      Intel isn't arbitrarily going, "man, we could make chips that run ok 5 degrees hotter, but we're gonna piss everyone off by demanding more cooling. Just because we can."

      this actually just the way the manufacturing process works. if the chips get binned at the top end - it gets the Extreme moniker. my personal processor is a xeon quad core - same exact chip as the q6600 - only my chip as a spec of 85*C while the consumer chips have a spec of 71*C in other words - my friend who bought the Q6600 - is the person intel is 'pissing' off because he has great cooling needs than i do.

      It's a zero-sum game. The total load on the planet is the same. The same total bunch of chips exits Intel's fabs. On the total, no energy was conserved.

      So Google's "going green" is at the cost of making everyone else less "green". They can willy-wave about how energy efficient they are, by simply dumping the difference on someone else.

      the total load on the planet is not the same. it is much cheaper to not cool an object than it is to cool it. other people's chips are not going to cost more 'green' points (or whatever is used to measure green friendliness [dont think carbon credits works here]). this is the functional equivalent of google saying 'i want only extreme processors' i simply do not understand your point of google dumping the eco burden on other people.

      lets not confuse hot CPUs with an affect on global warming or anything like that, because its is much cheaper (for the environment) to deal with the heat, then the heat + energy cost of the alternative. and when i say much cheaper, i mean order of magnitudes cheaper.

      It's akin to me willy-waving that I'm so green and produce less garbage than you... by dumping some of my garbage in random other people's garbage bins across the town. Yay, I'm so green now, you all can start worshipping me. Well, no, on the total the same amount of garbage being produced

      to be clear on what i mean, and to use your metaphor - it is not the same amount of garbage being produced since the air conditioners are being turn down. heat != pollution

    17. Re:But they pass it off to someone else by adpowers · · Score: 1

      Or you could read their power provisioning paper.

      Even though the individual machines aren't always fully utilized, in aggregate they can make assumptions about utilization and run very close to max at the rack and PDU levels.

      Also, I agree that this isn't necessarily a zero-sum game. If the chips have the same power profile but are more durable at higher temperatures, then they are better off at Google. Most other datacenters don't run as hot, so the extra heat durability the chips provide would be wasted.

    18. Re:But they pass it off to someone else by jonaskoelker · · Score: 1

      It's a zero-sum game. The total load on the planet is the same. The same total bunch of chips exits Intel's fabs. On the total, no energy was conserved.

      Let's see. Google runs their CPUs in a warm environment. Now they can turn the AC down a little. I run my CPU in a cold environment (Denmark, normal living room, not exactly perfectly insulated). The environment assists a bit in cooling, probably more than Google's environment does (even without their artificial cooling).

      I have to cool more than if Google didn't get the heat-tolerant CPUs, but Google can turn their cooling down more than I turn mine up, because my cooler environment already does some of the work.

    19. Re:But they pass it off to someone else by Anonymous Coward · · Score: 0

      Perhaps not, but it is fairly reasonable to believe that Google puts a much higher average load on their chips than the average Joe would. Hence, his point still stands.

    20. Re:But they pass it off to someone else by Mspangler · · Score: 1

      The obvious question is why Google didn't specify Mil-Spec chips if they wanted ones rated for more heat? Or does Intel not make them anymore?

  18. Re:WTF? Lawyers as engineers, not so much by kipman725 · · Score: 1

    chips near the centre of the wafer are higher quality. All google is asking for are these chips instead of a random mix of those from all over the wafer. This is why some chips over clock far better than others even though they were produced at the same week from the same plant. They can presumably ask for this because they are buying such large quantities. It's quite a novel way of saving money though.

  19. the obvious solution by sandmaninator · · Score: 0, Offtopic

    If they want the chips to run hotter, why not just use poorly conducting heat sinks?

  20. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  21. higher chip temps??? by sam_paris · · Score: 1

    I might be missing something here, but why would Google be demanding "higher" chip temps to save on cooling??

    Surely they should be demanding lower chip temps.. or is it just a mistake in the headline?

    1. Re:higher chip temps??? by OneSmartFellow · · Score: 4, Insightful

      They are not asking for the chips to be made to produce more heat, they're demanding that Intel guarantee that the chips will still perform, even if operated at a higher than specified max operating environment temperature.

      You would be forgiven for thinking it makes more sense to for Google to insist that the chips produce less heat, rather than that they will still operate in extreme temperatures, since the majority of the cooling cost come from dissipating the chip heat from the enclosed space. But hey, it's Google, they do things a bit different.

    2. Re:higher chip temps??? by NixieBunny · · Score: 2, Informative

      They don't want the chips to get hotter than they already do. They want them to work correctly when they are run hotter. This allows them to use passive cooling in more climates, which saves big bucks on the cooling bill.

      --
      The determined Real Programmer can write Fortran programs in any language.
    3. Re:higher chip temps??? by Anonymous Coward · · Score: 0

      they want chips that can survive under hotter conditions so they spend less on cooling - ie - they run the room at 80 degrees and know that the processor can handle the 80 degrees.

      typically, we would expect that you want a chip that runs at 65 degrees instead of 70 degrees while delivering the same performance - reducing total heat output.

      what Google is looking for is chips that perform hotter without failure, so they can cut cost on cooling their data centers.

      hope that helps...

    4. Re:higher chip temps??? by thered2001 · · Score: 1

      I think they mean chips capable of operating correctly at higher temperatures.

      --

      If your only tool is a hammer, every problem becomes a nail.

    5. Re:higher chip temps??? by sam_paris · · Score: 1

      Thanks both of you for putting me straight there. It makes complete sense now :) Although I still think the headline and summary could have been worded better...

    6. Re:higher chip temps??? by Tsunayoshi · · Score: 1

      They don't want chips that put out more heat, they wants chips capable of running in a higher ambient temperature environment. i.e. a server room maintained at 80 degrees Fahrenheit vs 70 degrees.

      --
      "Get a bicycle. You will not regret it, if you live." - Mark Twain, "Taming the Bicycle"
    7. Re:higher chip temps??? by sydneyfong · · Score: 1

      could have added the word "tolerance" somewhere...

      --
      Don't quote me on this.
  22. They should also consider... by Muad'Dave · · Score: 1

    ...using more effective means to extract the waste heat from the processors they already have. Lower thermal resistance equals lower operating temperatures. As many boxes as they have maybe they should invest in large-scale refrigerant-based cooling system with tiny heat exchangers for each CPU. I envision overhead refrigerant lines with taps coming down into each cabinet to serve the procs in it. Each server could have quick-disconnect lines on the back for easy removal. No need to cool all that air, and you'd get very good thermal resistance figures.

    --
    Tiller's Rule: Never use a word in written form that you've only heard and never read. You will end up looking foolish.
  23. Re:WTF? Lawyers as engineers, not so much by rrohbeck · · Score: 1

    Right on.
    Of course, Intel will give them whatever they want because Google is such a large customer. And will then pay in terms of higher failure rates, hence warranty costs. And Google will notice the same thing, assuming they do decent data gathering on failures, and find out that this is a really bad idea because those failures cost them even more than Intel.
    Seems to me like bean counters are trying to beat physics.

  24. Re:good. by CubicleView · · Score: 2, Insightful

    I wouldn't say the post is off topic. When I read it first, I thought the title suggested Google wanted the processors to produce more heat.

  25. Higher temp = higher power by hpa · · Score: 1

    I have to say I am a bit surprised. A CPU operating at a higher temperature will draw more power and thus produce more heat at the same performance point. This is one of many temperature dependencies in silicon circuits. Now, it's possible that Google's demand is that they can run at the same speed and power at the higher temperature, which means in reality they are underclocking a faster chip to run it warmer.

  26. Re:WTF? Lawyers as engineers, not so much by Alereon · · Score: 1

    There is very little Intel can do to stretch thing by another 5 degrees. It's not something that can be imposed by fiat. Intel engineers have already juggled all the variables to come up with the best performance possible. SOMETHING is going to have to give. Either the chips will have to be selected and graded for speed, lowering the overall envelope for the chips everyone else gets, or they'll have to fudge some other parameters, hoping nobody will notice, or worse yet they'll tweak some variable right to the edge of raggedness, resulting in worse reliability down the road.

    In the real world, processors don't fail (barring power spikes/motherboard failures that fry them). The consideration here is much more likely to involve legal concerns about the warranty or the temperature at which thermal throttling or shutdown occur. Most likely Google and Intel were both able to confirm that the processors would not fail during their expected lifetimes in Google's datacenters even when operating continuously at this new maximum load, which is why they agreed to amend the processor specifications. I sincerely doubt these CPUs are different from others in any way other than possible the thermal protection setpoints that are pre-configured.

  27. Re:But they pass it off to someone else - WRONG by MadCow42 · · Score: 4, Insightful

    >> So Google's "going green" is at the cost of making everyone else less "green". They can willy-wave about how energy efficient they are, by simply dumping the difference on someone else.

    The difference is that Google is going to actively exploit the ability of those hand-picked CPU's to run hotter. Chances are that the users who would have otherwise received those chips would not reap any energy savings from the capabilities.

    At a minimum, Google is contributing here by forcing a vendor to differentiate chips that have a capability of running hotter from ones that don't. No matter who uses that capability, it's a benefit to the planet (versus the alternatives at least).

    MadCow.

    --
    I used to have a sig, but I set it free and it never came back.
  28. But heat also affects... by invalid_account · · Score: 3, Interesting

    Hard disks. In fact, I am typically far more concerned with long-term issues with my data than with the computing itself. Not to mention, the CPU is NOT the only chip that can suffer from heat issues.

    1. Re:But heat also affects... by HardCase · · Score: 1

      Google doesn't use (or uses very few) hard disks. And who's to say that they have not asked for guarantees of performance from other chips?

    2. Re:But heat also affects... by NerveGas · · Score: 1

      Not only that, google did statistical analysis of a LOT of hard drives, and found that within a certain range, they actually lasted LONGER as temperatures went up.

      --
      Oh, you're not stuck, you're just unable to let go of the onion rings.
  29. Not necessarily about money/costs by Anonymous Coward · · Score: 0

    Alfred Spector (Google VP of Research & Special Initiatives) recently spoke at a recruiting event at my school. While going over some of the work he is in charge of, he mentioned that they had found that the optimal temperature to minimize hard disk failure was actually higher than what is generally accepted in most datacenters. Because of that, Google keeps their datacenters a few degrees warmer than most other companies do. Sounds like that's what this is about. Not necessarily saving on energy costs, but improving the reliability of their hard disks (and we know that Google has a huge interest in doing that)

  30. Higher operating temp by bperkins · · Score: 3, Informative

    There are two issues with higher operating temp.

    One is that you get less drive current from your transistors, so you get less performance (which everyone seems to understand), but this is usually a fairly small effect for 5 degree C.

    The _big_ deal with 5 degree C would be electromigration in interconnect metal, which goes up very quickly with temperature. So the difference in failure rates might be quite large.

    If there was any deal at all, it's likely that the Intel engineers tried to remove some conservatism from their temperature estimates to see if they could squeeze out 5 degrees from the thermal budget, or perhaps information on the workload itself to get Intel to "bless" the higher data center temperature.

  31. They should move data centers to Northern Canda by Twillerror · · Score: 1

    Then just leave the windows open.

  32. Re:Is this possible? I was going to say, "For such by davidsyes · · Score: 1

    a *hot* topic, Google raises a cool argument." But, because you remind us of the Arctic option, i'll have to say YOU raise a cool argument/reminder.

    But, is there a feasible way for various sloped shafts to be cored (or existing ones, such as the former Super Conducting Semi Collider (or is it Semi Conducting Super Collider?, etc...used) such that filtered draft air (sounds like beer, huh?) is blown past the chassis?

    And, isn't there a way to decouple the processors from such numerically high boards? Can't these processors be (to bring up images of the Star Trek USS Enterprise (NCC-1701 D) central computer core, or even the main warp core, with vertical shafting, but horizontal/azimuthal projections) attached to shafts, and the ancillary wiring be attached down/up stream? Then, the cooling air could be better directed, controlled and overall flow demands reduced, to in essence, cut the high energy costs.

    For visuals, see:

    http://startrekspace.blogspot.com/2007/01/geordi-la-forge-and-his-warp-core.html

    http://www.loony-archivist.com/lowerdecks/life.html

    http://www.ussenterprise.co.uk/enterprise/entd/

    I would envision that at the very least, Google can -- or already has underway -- plans to exploit polar or Canadian, or cold North Dakota type environs in which to shaft-locate their computers.

    --
    Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
  33. Higher temperature is overrated by Skapare · · Score: 1

    Temperature is just the stable point across a heat flow. You could immerse an entire data center in liquid nitrogen, and you still only need to remove energy at the same rate in goes in. One problem with a lower temperature (relative to outside) is that the rate of leakage from outside is higher. That's fixed by having more/better insulation, in addition to raised solar panels covering the entire building.

    Higher temperatures also increases resistance in conductors and electrical parts. That leads to more I2R power loses which is energy that doesn't get used but still has to be removed. Superconductors would be a plus for some of the core power feeds, too (you still do have to cool them).

    Computers should be fed power at the highest practical voltage to keep current low (I2R losses, again). Most power supplies are more efficient at the lower current levels. That means at least 230 volts, if not higher. Commodity power supplies handle that fine. It wouldn't take much design change to bring them up to 277 volts (US) or 347 volts (Canada). And of course there might be even better efficiency involved in delivering 340 volts DC right to the PSU designed for it (most AC ones have that voltage as an intermediate step, anyway).

    And efficient coding can be a factor here, as well, at the scale Google operates. Fewer cycles used for a given transaction means less energy needed to carry it out. Then the CPU goes back to idle and the clock can be kicked back to low frequency. And fewer machines needed to do searching means more machines can be completely off.

    --
    now we need to go OSS in diesel cars
    1. Re:Higher temperature is overrated by Sebastopol · · Score: 1

      if you design a part that can run at a higher temperature, you save money in cooling. just a few degrees C can translate to millions of dollars in a server farm.

      --
      https://www.accountkiller.com/removal-requested
    2. Re:Higher temperature is overrated by Skapare · · Score: 1

      There's more than just design involved in parts that can run at higher temperatures. These parts will also be more expensive to manufacture. You also need to design and manufacture the other parts to operate at the higher temperatures. And you also need to make sure all the parts operate more efficiently at higher temperatures (normally as temperature goes up, parts are less efficient). Compare that to the cost of better insulation for the building. Compare that to the cost of arranging the building construction so that there is less surface for thermal ingress. It is a big complex model to work out.

      --
      now we need to go OSS in diesel cars
    3. Re:Higher temperature is overrated by Sebastopol · · Score: 1

      here's more than just design involved in parts that can run at higher temperatures. These parts will also be more expensive to manufacture.

      I don't follow. Outside of the CPU, what parts cost that much more to manufacture to run at 5C higher? Motherboard? Fans? Chassis? Power Supplies? AC for the server closet? The cost of running the AC will increase, but I think 5C is only a big deal on the die, and maybe for acoustic reasons in the case ambient temp.

      I was under the impression the extra 5C in the CPU was simply an increase in the thermal gradient used to spec the parts as they approach catastrophic conditions. The amount of silicon operating at the extra temperature introduces a near nil extra (Q) energy into the chassis, since we're only dealing with the hotspot here, but it does have a huge impact on the case to heatsink properties which increases the heatsink cost.

      Outside of that local blade/chassis impact, the cost should be very small.

      Fill me in as to what I'm missing here.

      --
      https://www.accountkiller.com/removal-requested
  34. Google Demands that the Water Authority by netglen · · Score: 2, Funny

    to supply Google with water that boils at 5 degree C less. That way Google will save millions of dollars on costs of making coffee & tea for their employees.

    1. Re:Google Demands that the Water Authority by Skapare · · Score: 1

      That's what they deliver in Denver.

      --
      now we need to go OSS in diesel cars
  35. Better Cooling or change platform by Anonymous Coward · · Score: 0

    Switch to AMD, the servers run on lower power RB-DDR2 not FB-DDR2 and the chipsets are much cooler.
    Replace the CPU coolers with more efficient TI to AIR coefficitient sinks thus allowing a warmer room temp as the sink is able to move heat more efficiently. Opterons also take advantage of load distribution moving the running processes from core to core allowing overall cores to remain cooler yet with warmer room temps.

  36. It's not a zero sum game by S-100 · · Score: 1

    It's a zero sum game if Intel does nothing within its capabilities to shift its production to higher temperature chips. Since they are getting paid a premium to provide them, they are motivated to tweak their processes to make all their CPUs more efficient. And if that's the case, your premise is completely wrong.

  37. davey jones' AC by bugs2squash · · Score: 2, Funny

    Isn't Google close enough to the ocean to pump cold water out of the depths to help pre- or post- cool air ?

    --
    Nullius in verba
  38. Re:WTF? Lawyers as engineers, not so much by HardCase · · Score: 1

    There is very little Intel can do to stretch thing by another 5 degrees.

    They'll either agree to the guarantee with their standard chips or they'll bin the chips just like they do with speed grades. My company does the same thing with the chips that it produces. We have a commercial temperature range and an industrial temperature range - they're the same chip, but some are binned for the higher temperature.

  39. Re:good. by OrangeTide · · Score: 3, Insightful

    What? American businesses like saving money almost as much as they like making it. It's environmentalism is not as a big motivator as profit, at least in the US. Make being efficient profitable long term, and some businesses will do it. Make it profitable short term and businesses will fall over one another to do it.

    --
    “Common sense is not so common.” — Voltaire
  40. ...in other news! by chaney · · Score: 0

    Intel demands that Google no longer require any sort of internet connectivity for the use of their search engine, as Intel spends hundreds of thousands a year in bandwidth costs.

  41. Project 02 by djupedal · · Score: 1

    >One way around it could be to locate datacenters at locations with natural cooling available like rivers and larger lakes.

    Google has a large data center on the bank of the Columbia River in Oregon now...

    1. Re:Project 02 by Alpha830RulZ · · Score: 1

      The primary reason for this is access to power from the John Day Dam, and a site that is removed from seismic and weather issues.

      --
      I was taught to respect my elders. The trouble is, it's getting harder and harder to find some.
  42. Re:WTF? Lawyers as engineers, not so much by Alzheimers · · Score: 1

    It sounds like what Google's looking for isn't so much a technical change as a contractual one.

    They basically want Intel to replace burnt-out CPU s that are run over spec. Chips are usually spec'd lower than what they're physically capable of (see also: Overclocking) to extend their MTTF to a point just beyond their warranty.

    From the article:
    [i]If the chips failed prematurely at these higher temperatures, the former Googler says, Intel was obliged to replace them at no extra charge.[/i]

    CPUs aren't the only point of failure in an over-hot system. Thanks to recent reformulations of solder, a lot of the contact points on modern electronics easily separate at lower temperatures (see also: Xbox360 RROD)

  43. Svalbard by andersh · · Score: 1

    Actually you could easily do this on the Norwegian Svalbard islands. Not only does it have Arctic conditions, reliable power supplies and high-speed fiber connections to main land Norway/Europe. But it's also a special status island group where citizens of any nation are free to live and work. Being part of Norway it's under a stable, free and democratic government.

    Because of it's position close to the North Pole it's heavily used as a satellite communications site by NASA, JAXA and the European Space Agency.

    There's even a University there.

    Svalbard Satellite station (SvalSat) was established in 1997 and the rapid expansion of the ground station is changing this perspective. SvalSat is recognized, not only as the northernmost, but also the best-located ground station in the world. The extreme northern location on the Svalbard archipelago, at 78Â13' N, gives SvalSat its unique and favourable position. The satellite coverage at this latitude holds unique opportunities and SvalSat is the only commercial ground station in the world able to provide all-orbit-support (14 of 14 orbits) to owners and operators of polar orbiting satellites.

  44. Proper naming by Anonymous Coward · · Score: 0

    I thought everyone knew we don't say centigrade anymore. It's Celsius since 1948, sheesh.

    AC

    ps: captcha was appropriately "apology"

  45. After dinner by Unsung+Bovine+Herd · · Score: 1

    If it were that warm, those bears would be doing something else.

  46. It's called 'screening' by pbrooks100 · · Score: 1

    This is called 'screening' and is common in the electronics components industry. Your 2.66 GHz CPU is the same as a 2.8 GHz CPU but during screening started to generate errors at 2.8 and was simply sold as a 2.66

    Google is paying (or insisting) that all their processors to be screened at +5 degC. They may pay a couple of dollars more for this, or may hold the fear of using AMD processors over Intel's head to get this testing at no additional charge.

    Screening is done in the simplest technologies including resistors. The difference in a 1% and 0.01% 100 Ohm resistor is the price paid for screening, not necessarily the quality of the component...

  47. Call me crazy... by spectecjr · · Score: 1

    But why don't they just use mil-spec grade chips? Aren't they already graded for -55 to +125C operation. They're usually radiation hardened too.

    --
    Coming soon - pyrogyra
    1. Re:Call me crazy... by CAIMLAS · · Score: 1

      Ok: you're crazy.

      Milspec stuff always costs more. ALWAYS. If Google is trying to cut costs, spending more isn't the solution.

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
    2. Re:Call me crazy... by spectecjr · · Score: 1

      But what they're asking for is basically cut-price Milspec. *shrugs*.

      --
      Coming soon - pyrogyra
  48. So one former employee says something... by Anonymous Coward · · Score: 0

    ...so it must be true...interesting...

  49. What about cooler chips by TechwoIf · · Score: 1

    Why higher temps? Google should be demanding chips that run cooler. Cooler running chips save more money then running a data center at a higher temp.

  50. Re:WTF? Lawyers as engineers, not so much by paulgrant · · Score: 1

    if you were an engineer, you would realize they optimize all the variables including cost of production -- who's to say they made a conscious decision not to run hotter simply because they would have to charge X amount of dollars more per chip? For a company specializing in data-centers, trading in a variable cost for a fixed cost, its a win-win situation ;)

  51. Re:good. by Migity · · Score: 0, Flamebait

    When are you going to stop being such a troll?

  52. WTF? by CSMatt · · Score: 1

    OK, could someone please explain to me how running hotter processors saves money on cooling costs? It seems that it would actually increase cooling costs.

    1. Re:WTF? by Anonymous Coward · · Score: 0

      It's easier to cool things when they're hot. Suppose you have two rods at 100 and 50 degrees in a wind at 40 degrees, which of the two rods will lose 1 degree faster?

    2. Re:WTF? by SagSaw · · Score: 1

      OK, could someone please explain to me how running hotter processors saves money on cooling costs? It seems that it would actually increase cooling costs.

      If the processors can operate 5 degrees warmer, the room can be kept 5 degrees warmer while the processors do the same amount of work. This should cut down on cooling costs.

      --
      Come test your mettle in the world of Alter Aeon!
    3. Re:WTF? by CSMatt · · Score: 1

      But wouldn't this necessitate that the central air would have to run more often to cool the processors during the summer, canceling out any overall cost savings? It may be that Google's servers ultimately use more juice than their air conditioning, but there's still the nagging problem that a lot more of the electrical energy is dissipated as heat, and to me that sounds very inefficient.

  53. Re:good. by raidfibre · · Score: 1

    Higher temp does not equal more heat. Just thought you should know that.

  54. Sounds good by Anonymous Coward · · Score: 0

    I think we should support Google's quest to have hotter servers so they can harness the heat output to power the hover craft engines for their floating data centre ... on the sea ... of the moon ... of Neptune ... in a parallel universe. Why would Google want to do this? Because they can. Why should we support them? Because that would be neat.

  55. Degree Celcius!!! by Anonymous Coward · · Score: 0

    Centigrade is one hundredth of a grade which is an angle

  56. Not everyone is in Denmark, ya know? by Moraelin · · Score: 1

    Not everyone is in Denmark and not all of Google's servers are in the Silicon Valley, ya know? Equally, now some of those chips which need less cooling will go to some Google server in Ireland (check it out, they're about as high up as you are on the map), and some which need more cooling will go to some poor sods in Mexico or Saudi Arabia or Israel or God knows which other hot place.

    --
    A polar bear is a cartesian bear after a coordinate transform.
  57. Centigrade by Barryke · · Score: 1

    Finaly, thank you! Centigrade!

    No more Fahrenheit in frontpage articles please.

    --
    Hivemind harvest in progress..
  58. Re:WTF? Lawyers as engineers, not so much by truesaer · · Score: 1

    I don't think you've got the full understanding. The envelope for a product has to take into account the fact that they need to be able to produce that product in large volumes. There will be many parts that test out as being able to operate at a higher temperature, but perhaps not enough to justify an entire new model of the processor. So to fulfill an order from a single customer for parts rated to a higher temp, all they have to do is set aside some chips that exceed the standard product specs. These parts always exist, you just never have any idea if you've got one or not. Google apparently had the clout to basically get a special processor model available just to them.

  59. Re:Won't that heat up the arctic by Douglas+Goodall · · Score: 1

    If we put all those hot chips up there, we will melt the snow faster and them people will complain. The real trick would be to take the heat from the chips and use it to generate energy to run the chips. Sort of a perpetual motion deal.

  60. Re:good. by CubicleView · · Score: 1

    I do, and I implied as much by admitting to my initial confusion with the title and specifically using the word heat.