Slashdot Mirror
Using Jet Engines to Cool Servers
rpmsci writes "The computer servers that fill huge data centers are producing more heat with every new generation of processors. It's a problem that's sending engineers on a search for cooling fans that are both small enough to fit inside ever-smaller server chassis and powerful enough to dispel increasing amounts of heat. At Hewlett-Packard, they've found one answer in an unexpected place: model jet airplanes."
ASUS beat them to it
Javascript + Nintendo DSi = DSiCade
These are just ducted fans. There are actually tiny gas turbine engines available for model aircraft.
I have to wonder how much if this is really just hype. Last time I looked at my cooling fan it was already a ducted fan.
Are they adding extra stages? Maybe more an more efficient airfoil on the fan blades? Longer duct? Higher RPM?
I find this a huge so what.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Christ, is "active" a hip marketing term again? I thought "ActiveX" put a bullet in that fad...
Air is such a poor heat transfer medium. Why not build a rack with a water cooling system built in? I have an external water cooled solution on my home PC connected via a set of no-break quick release couplings. So any time I need to pull my PC apart I can pop the coolant lines with out losing a drop of coolant or introducing air into the system.
I can't imaging running a fleet of model airplane engines is going to be quite, cheap, or all that reliable. Especially when compared to an rack integrated water cooling system.
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
http://www.asciimation.co.nz/beer/
Yeah, real innovative HP. *yawn*
Sheesh, a ducted fan is NOT a jet engine.
It just means they put a fan inside a tube, rather than have a propeller outside.
It's a cosmetic thing to keep the appearance of a jet.
Other than the fan being in a aerodynamic tube, it really isn't any different.
That asus is just a standard fan mounted in a case that looks like a jet engine, but it's the same technology.
/. but sometimes it's worth reading.
On the other hand, the HP one uses small blades that are shorter and that spin faster. As such they create more thrust/airflow and reduce noise that normal blades produce from the tips of their blades.
RTFA, it's got a good discription, yeah, I know it's
I am SO disappointed. I was hoping it would be the real thing.
That would've been entertaining.
You might want to reinforce the footings for those racks before you.. *WHINE* *ROAR* *CRASH*
--
silas
Those 400psi water chillers on IBM TCM (Themo-Conduction-Module) mainframes did an awesome job of keeping them cool. I'm sure some research into it could yield smaller cheaper units with better thermal conductivity - perhaps some new fluid that's non electrical conducting and better heat characteristics.
They are not using Jet Engines, they are adpating Electric Ducted Fans, which while they operate on some of the same principles, are very different things. (No combustion chamber, no hot exhaust)
1) Tech installs HP jet fans into all his servers.
2) Tech restarts all his servers.
3) Tech shits himself as all servers simultainiously (sic) take flight smash though the wall and put themselves into a flight path heading fot the North Pole.
To err is human. To forgive is not company policy.
Many such problems are due to the need for the massive amounts of hardware necessary to run many enterprise-grade software packages. Java (and .NET, as well) is one of the most significant causes of such problems. Because of its architecture, the amount of processing necessary per "task" (as in a set of actions, not as in executing processes/threads) is vastly overinflated.
Simply put, software like Java and Oracle isn't a very effective user of CPU resources, thus necessitating the need for vastly more CPUs to be used to perform even basic functionality. I know, I know. We've all seen the benchmarks where Java outperforms some C or C++ code in some limited benchmark, and that JDK 1.5 or the prereleases of 1.6 are far more performant. However, real-world experience tells us that Java consumes far more computing resources than equivalent programs in other languages.
Rewriting a large, enterprise Java app in C++ (or for safety, even a language like SML) can often lead to the consumption of far fewer computing resources. Whereas a certain Java application may need a cluster of 90 to 100 2.8 GHz Opteron systems to run effectively, a similarly-written (ie. algorithms of the same complexity, etc.) system in C++ can make do on 15 of those machines. Using only 1/6 or so of the machines necessary does cut down on the power consumption significantly. That directly leads to financial savings for all of the companies and individuals involved.
Apple did this years ago, but with real jet engines. Its called the PowerMac G4. I was so impressed by this advance, that I immortalised it in my sig.
Nothing sucks like a Vax, nothing blows like a PowerMac G4
What about the guy who submerged his entire computer in cooking oil? (Or was it mineral oil?)
Maybe they should consider using the thrust after all. Wasn't there some mention, a few months back, of the "world's smallest flying web server?"
Better yet, use a real jet engine. (Better make it suck air out rather than blast air in, though.) Bonus: hook it to a water pump for the water-cooled machines. (They really do have gas turbine water pumps, btw, used in fire-fighting. Portable and powerful.)
Intel already killed netburs.
Sheesh, Zonk -- could we at least take, say, three seconds to think before writing the article title. How about "Using jet engine technology..." instead of "Using jet engines..."
_ things_work/default.jsp) sucking JP4 and blowing 1000's of cubic feet per second of very hot air into the server room here at work.
Little clue: Jet exhaust is... well, let's just call it "a little warm for cooling a server" and leave it at that. The article title gave me this picture of a Rolls jet engine (http://www.rolls-royce.com/education/schools/how
Oh the humanity!
Interested in a Flash-based MAME front end? Visit mame.danzbb.com
From TFA:
"They literally blow you away," he says; "it's like picking up a leaf blower."
Great. As if I don't get enough of that sound when I'm trying to sleep in on weekends...
Now, when people come over and my computer is humming along and they ask me "what the hell, do you have a jet engine in there?" I can look right at them with a smirk and say, "why yes, I do".
Arguing with an engineer is like wrestling a pig in mud. Soon, you realize the pig is dirty, and he likes it.
Use jet engines to cool servers?! Are you mad? Can you imagine how much this will ramp up global warming? Is it really worth it?... Yeah I guess so. Nevermind
Support the FairTax
For a moment I thought they were using actual jet engines to cool the server, but noooooo, they had to go for boring ol' electric fans instead.
[insert rant about misleading summary]
--troll;
They couldn't hear the complaint that "your hardware sucks and blows at the same time" so they had to come up with something so they can just smile and nod.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
What I don't get is this: Since there is all that heat coming off of processors, etc...how can the energy in that hot air be captured and put to good use? Seems all cooling solutions just "consume" more energy to transfer energy (in the form of heat) from here to there.
The net effect is we take a bunch of energy (as electricity) and lose a lot of it as heat then take a bunch more of it (again, electricity) to just move the other "lost" energy (heat) around. It just seems wasteful and expensive to me. There's got to be a better way; not that I could tell you what it is.
(note: I put quotes areound "consume" and "lost" because I understand the energy isn't really lost or consumed...but that's sort of what it's like and I'm at a loss for better wording.)
If a jet engine can cool a can of beer, why not a server?
That's a hard lesson learned there, spend the $35 for a non-conductive liquid and save hundreds, if not thousands in hardware costs. The same thing applies to UPSes.
The cesspool just got a check and balance.
One could compromise by making the units hermetically sealed, and piping some other gas than air that conducts heat better. Less worry about fatal leaks.
Of course, HP is modifying the fan design to optimize the pressure which is apparently different from thrust.
It makes sense that this would be the case: if you think about how a jet engine is mounted, there's no backpressure on the exhaust stream aside from atmospheric pressure, which tends to be constant at the altitudes that RC planes fly. However, if you were to mount it so that the out-flow was restricted (because it's blowing into a computer chassis), then you'd need to redesign the blades. My initial guess is that the pitch has to change, although I haven't done any analysis of it.
What I wonder is whether the pitch of the blades and the overall design has to be changed for each application? Depending on the configuration of the machine these fans are going to be used in, the back-pressure (outflow restriction) is going to be different, necessitating different design optimizations.
It would be interesting if they produced a fan that had variable pitch blades, controlled by some sort of servo, that would maximize flow or pressure in different situations. Perhaps such a system could even be used to regulate airflow while keeping the spindle turning at some fixed speed at which the motor was most efficent.
I don't do ducted-fan RC stuff (helis are my thing, personally) so I don't know if anyone has ever made a variable-pitch miniature jet engine. If they did, then the HP guys might be able to use that without much modification. Even if nobody has done it before, it doesn't seem like it would exactly be that difficult, with the correct budget and resources available (like a good shop and skilled prototyping machinist).
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Powermac G5
They use Delco pumps and radiators in each CPU module.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Since these ducted fan's aren't really jet engines, and certainly aren't what I had in mind when I saw the term "jet engine" in the headline (think very large and noisy!), here's a proposal for using a real, full-sized, jet engine for cooling your servers:
Take one jet engine,
Add stages to capture the thrust and transform it into more torque,
Connect output shaft to massive freakin' compressor turbine,
Use turbine to compress gaseous coolant back to a liquid,
Attach big large radiator/heat exchanger/water cooling tower
Viola! you now have many tonnes of refrigeration capacity, good for blowing cold air through your equipment room, or circulating liquid coolant directly to the chips.
The best part is, you get to have a jet engine tacked on to your server farm.
Actually most datacenters already have massive water-cooling systems: only they're building wide and generally used to cool the air. I'm talking about the HVAC system, of course.
Large buildings generally don't circulate Freon from one floor to another, it would be too expensive. Instead, they have a big refrigeration unit (roof mounted, usually) with big cooling towers and the rest, and use it to chill water, which is pumped throughout the building and used to cool air.
It wouldn't be very difficult to tap into the chilled-water lines that already exist in most buildings, and use them to cool the servers directly. In fact this was once a lot more common: back in the day, it wasn't uncommon for big mainframes to be water-cooled. I've worked with big scientific apparatus that's also water-cooled, and a lot of it used lots of electricity as well, so it's not as though the engineering is impossible.
Yes, there are certain risks associated with having water flow through your computer system, especially in regards to leaks. But there are lots of pieces of equipment that contain liquid and wouldn't appreciate leaks, and we don't think twice about them. For very valuable systems, an additional cooling loop filled with a non-conductive (or even better, a pressurized gas) coolant could be used, with a heat exchanger connecting to the building chilled water.
I think there are some IBM blade systems out there right now that use liquid cooling, but for some dumb reason they won't accept building chilled-water connections (believe it or not, they need water that's warmer than most building supplies). I can't find a link to it right now, but basically it introduces an additional heat exchanger for the sole purpose of warming the incoming chilled water supply before circulating it through the systems. Obviously, this limits their attractiveness and ease of installation.
But at any rate, I think going to liquid cooling, whether water or glycol or something else, is eventually inevitable in high-density applications: despite some of the practical problems involved, when you look at the economics, cooling is one of those things that scales really well. It's going to be cheaper in the long run to reduce the number of heat-transfer steps in between the chip and the outside environment (where the heat is going one way or another), and to do it all at once if possible. Maybe we haven't hit the power/density break-even point yet, but we must be getting close.
I think the reason you still see a lot of air-cooling is because the mass-production of components has made it inexpensive to do, but blade server systems are starting to run into the limits of commodity hardware (as this whole story with HP's fan attests to). When you start having to consider developing specialized cooling hardware anyway, whether for air or liquid, suddenly liquid cooling becomes more attractive.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Won't it then become conductive, or rather the mixture of wet dust?
Infuriate left and right
That's about all the article says.
The key ingredient to a ducted fan is efficient expansion. Any old array of twisted parts can propell air. I read another article and fabricated such a thing from Dixie cups. After your rotor comes the stator, a very important component missing from ordinary fans, which removes the angular component of the flow velocity. You want to move the air down your axis not around it. Getting the air moving along the axis and expanding it out to larger volumes without wasting your effort is hard to do. Adding any stator will help. Doing it quietly and efficiently is one of those rocket science things.
Wikipedia, of course, has a quick article,
and Google turns up an easy design text.Friends don't help friends install M$ junk.
In the article it mentions that this will help reduce the need for improved air conditioning in the server room (or something like that), but I can't help but wonder how they expect to achieve efficient cooling if all you are doing is moving the same air arround. After all, any air cooling simply moves the heat from one place to another, so when the room's ambient temperature increases as a result of more rapid heat tranfer from the hotter and more dense collection of processors and other heat sources then you no loger have cool air intake for these advanced ducted fans to use.
:)
Not only that but I think I will be investing in companies that make earplugs; Look out construction industry, the IT department's demand will bring your supply to it's knees
- James
I thought that using JET would create more heat, rather than reduce it.
I saw an experiment that proved distilled water does indeed short out electronics. They dropped three separate running tv's in ordinary water, distilled water and cooking oil. Only the cooking oil one kept working, the other two shorted out and made fireworks as soon as they hit the water.
Those editors sure have an interesting profile of their target readers!
In a smaller, more modern environment, the effect on other computers is going to depend on whether the cases let water leak from one to the next - e.g. in a stack of 1U, are there vents on the top and bottom or only sides - and on whether there's enough liquid leaking to get to the other machine before it evaporates.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
How do you think jet airliners provide a/c? They use bleed air from the turbine, that's how. If you take a very high pressure gas and allow it to expand very quickly it cools down a great deal. The same principle could be applied to cooling a data center. Why you would want to, well that's why I read the article. Only to be disappointed by the lack of real jet engines.
Coding with assembly is like playing with Legos. Coding an application in assembly is like building a car with Legos.
As an HP Tech in their manufacturing division, I can tell you this, the fans do sound like a jet engine. But they also move a lot of air. And their on the back of a special box, with five fans each on top and bottom, so air flow is from the front middle through the back top and bottom. The blades do stay cool. And I don't know if they intend on utilizing this idea on the smaller fans as well, for now, they are only with the blades
2^3 * 31 * 647
Anyone everthought to pump said waste heat round a building to either heat water/rooms/divert to the handryers in the loos(or bathrooms. Although I have never had a bath at work whay are they called bathrooms when there is no bath in them!?)?
In the not too distant future, next Sunday A.D.