CoolIT Vs. Cogage: Little Water And Big Air Compared

By published

Test Settings

We overclocked our C0-stepping Core i7 920 to 3.80 GHz in order to assure a high thermal load with both high-performance cooling solutions, using a 200 MHz base clock and 19x core multiplier.

Test System Configuration
CPUIntel Core i7 920 (2.66 GHz, 8.0 MB Cache) Overclocked to 3.80 GHz (19x 200 MHz), 1.366V
CPU Cooler, AirCogage True Spirit at continuous 12V 120 mm fan, 1,600 RPM
CPU Cooler, LiquidCoolIT Domino A.L.C. at continuous 12V 120 mm fan, 600-2,800 RPM
MotherboardAsus P6T Rev. 1.01G, BIOS 0306 (02-03-2009) LGA-1366, X58/ICH10R
RAMKingston KHX16000D3ULT1K3/6GX (6.0 GB) DDR3-2000 at DDR3-1600 CAS 8-8-8-24
GraphicsZotac GeForce GTX260² 896MB 576/999MHz GPU/Shader, GDDR3-2484
Hard DrivesWD VelociRaptor WD30000HLFS 0.3 TB, 10,000 RPM, 16 MB Cache
SoundIntegrated HD Audio
NetworkIntegrated Gigabit Networking
PowerCooler Master RS850-EMBA 850W, ATX12V v2.2, EPS12V
OpticalLite-On LH-20A1L, 20X DVD±R
Software
OSMicrosoft Windows Vista Ultimate x64 SP1
GraphicsNvidia GeForce 182.08
ChipsetIntel INF 9.1.0.1007

 
A 1.366 volt output under full load (1.344 volt idle) resulted from setting our Asus P6T motherboard to 1.350V core with Load-Line Calibration enabled.

The increased voltage kept our CPU stable at 3.80 GHz throughout its 8-thread Prime 95 64-bit torture test, while pushing temperatures on these high-capacity coolers beyond 75 degrees Celsius.

Benchmark Configuration
Prime95 v.25.864-bit executable, Small FFT's, 8-threads
RealTemp 2.70Highest core reading at full CPU load (60 minutes) Highest core reading at 30 minutes idle
Acu-Rite Thermal ProbeIOH/VRM Heat Pipe temperature at northbridge junction Ambient temperature at 6" from motherboard front edge
Thomas Soderstrom
Thomas Soderstrom
Thomas Soderstrom is a Senior Staff Editor at Tom's Hardware US. He tests and reviews cases, cooling, memory and motherboards.
55 Comments Comment from the forums
  • Proximon
    I would have liked a better description of this Cogage. HDT? Aluminum block? Copper? It could have used a few more pictures as well.
    Reply
  • dragonsprayer
    Hey guys! of coarse....THG rocks... THG is #1 in tech! The concept of comparing air, water and tec/water is great! Unfortunately this is no easy matter! WE, WarpedSystems, KEEP THIS STUFF AS TRADE SECRETS! WarpedSystems build triple loop systems with 250 l/h and with 750l/h pumps. We build the WarpedSystems Hybrid Single Loop water cooling - which can integrate tech cooler such as the Boreas from coolit - sure i slammed cool-it and they deserved years ago - but they continue to improve. (Warped Hybrid Single loop = large tubes running of the pump or pumps then slip off after 1 or 2 rads and then addtional cooling on raditors 3 or 4) Sure, they did ship product that did not work (yes?)- try a qx6700 at 3.6ghz with the freeze> We did at great exoence and time and you needed a 5000rpm fan too cool the tec's. Sure, or yes i am willing to forgive them and try the new stuff and still try incorporate products like the borealis but pure water is the way to go for most! Ya dual phase coolers are cool but..... we our shipping 24GB ram, 4.5ghz 965.975 i7 systems that do not need tec's, phase coolers or what ever! I will close by saying, the i7 GREATLY BENFITS FROM WATER COOLING!!!!! BIG TIME! I will tell THG people we use hybrid single loop (on reserve and 2 pumps that split off after the intial heat removal, the we use cool air rad too feed the cpu, you want one save up $6k+! thx! THG rocks!
    Reply
  • JeBuSBrian
    Come on, Tom's! Hire a proof-reader already. Your writers obviously can't do the jobs themselves. And it simply makes this site look more unprofessional than it already has become.
    Reply
  • renozi
    Is the Spirit better than the TRUE 120?
    Reply
  • Crashman
    renoziIs the Spirit better than the TRUE 120?
    The article said they never sent the Ultra 120 Extreme.
    Reply
  • dragonsprayer
    lol! me? sorry about the typos - i am testing a system while i test this - keep in mind its running at 4.15ghz at 72c with prime 95 and 3d mark06 running at the same time!!!!! with hypo treading on! try that cyber-loser! or ibuyjunk! o man i should not do that sorry THG! i know they support the site but, when they claim 4.1ghz but do not spec 100c with prime 95 - well we get a little up set over here! hybrid single (keep in mind this is our low end system!) (includes push pull fans in 3 rads- look for the split at the upper left)anyone that wants to build ther own email me or pm me! tech support is free for THG, loop: http://s329.photobucket.com/albums/l373/warpedsystems/?action=view&current=WSZWC-cms-i7-2.jpg ,,,, http://s329.photobucket.com/albums/l373/warpedsystems/?action=view&current=DSC_0166.jpg enjoy and good luck! we do plumb the boreas in this too - sorry for no proof rading~!
    Reply
  • "At full speed (2,800 RPM) its fan was over twice as loud as that of the True Spirit."

    That will lead people to believe it was 56dB, when in fact something that is "twice as loud" is only +6dB
    Reply
  • Crashman
    Drikanna"At full speed (2,800 RPM) its fan was over twice as loud as that of the True Spirit."That will lead people to believe it was 56dB, when in fact something that is "twice as loud" is only +6dB
    No. Twice as loud is +10db. Twice as many fans would be +3db. We didn't have a accoustically dampened room to test them in, and taking meter readings against background noise provides skewed numbers that most people can't work out. It's easier to simply say "at least twice as loud" when you mean more than 10db louder.

    What makes that observation most significant is that both companies rate their devices at similar full-speed noise levels, when it's quite obvious to testers that one device is very much louder than the other.

    There are many ways to screw up a reading even in a silent room. The standard from audio devices is to test at one meter, but some review sites do it at around one decimeter. If you go in the other direction, perhaps the "noisier" device was tested at 1 dekameter.

    Any company who advertises noise readings that were taken at more than one meter without specifying the distance is cheating its customers. Any company who advertises noise readings that were taken at less than one meter is cheating itself.
    Reply
  • dragonsprayer
    rating and real life are very different, there are many factors in actual "feel" of the sound. I mean by "feel" is how loud somthing sound relative somenting else.
    Reply
  • dragonsprayer
    The photos i posted above - the system is running at 4.1-4.3ghz the range is due to turbo. I have been running 3dmark over and over with quad sli set up and the i7 had not past 80c this is a single loop. OK, its hybrid, we slit the loop atfter so many rads and then send the rest to reverse flow rad - air blowing in so the rad is cooled with cold air. OK THIS IS NOT AN AD! had any one else seen numbers even close to that? we running prime 95 in 64bit vista, with 3dmark 06 and getting less then 80c with hyper threading on. The 3dmark drops from 24-25k quad sli with 9800GX2 to 12-13k while running prime 95~ Dream it! build it! (screw those guys who say you dream it we build it! you dream it you build it! free help!)
    Reply