Noctua NH-D15 vs Corsair H100i V2 my comparison and my little test

Discussion in 'Die-hard Overclocking & Case Modifications' started by jura11, Sep 3, 2016.

  1. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    Hi guys

    Few weeks ago,my Corsair H100i V2 started to develop sudden increase of temperatures which has point me to,this can be down to the air lock in system which shouldn't happen or is down to the pump or simply system is clogged...One day I went to BIOS to check temps and saw 103.5C in BIOS,I turned down the system and replaced paste and then again tried few others thermal pastes like Thermal Grizzly Kryonaut,PK,Shin Etsu and AS5,still same,one thing what helped is bashing with hand where are end tanks on radiator or where are hoses go etc,with that my temps dropped to previous temps,but still temps always has been on higher side or on hotter side


    Then I decided,I've sent my Corsair H100i V2 to RMA and ask my brother if he can buy me Cryorig R1 Universal or Noctua NH-D15,he got Cryorig,but he asked in shop if its compatible with X99,they didn't know that and he returned back Cryorig which I would rather have,he bought me Noctua NH-D15

    Fitting of the Noctua has been very easy and I didn't expected fitting of this cooler can be any easier,previously have owned Thermalright and their mounting is strange and I've personally struggled :rolleyes: ,regarding the Noctua,fitting the D15 has been much more easier than fitting of the H100

    Here is my specs what I'm running and ambient temp has been same in both tests(I've done 2 hour render in 3DS MAX) around 18C

    ASRock X99 Extreme6
    i7-5820k OC 4.4GHz vCore 1.226v
    Thermal grizzly Kryonaut has been used in both tests
    GTX Titan X with EVGA Hybrid AIO
    Phanteks Enthoo primo with stock fans
    All HDD cages are full or loaded
    Corsair AX760i
    Fans controlled through the Phobya fan controller
    Ambient temperature: 18C



    Here is screenshot of SIV64 of my test with H100i with BeQuiet Pure wings 2 fans running in 1200RPM

    [​IMG]


    and here is screenshot from today my test with NH-D15

    [​IMG]

    As you can see I gained around 8C on load at PKG and on cores I've gained up to 10C which I think is pretty much awesome,looks like finally I can push my 5820k to 4.6GHz pretty much easy and I mainly finally I can sleep bit better and don't be worried about the temps,in games temps never goes above 52C on PKG,cores I don't monitor when I game,with H100i V2 PKG temperature has been in low 60's

    Not sure if this does help to anyone,but from my point of view I would say,if someone is still on CLC or AIO,my recommendation is ditch that and get nice air cooler,you will be surprised how it perform and how quiet operation can be


    Hope this helps guys

    Thanks,Jura
     
  2. tsunami231

    tsunami231 Ancient Guru

    Messages:
    14,750
    Likes Received:
    1,868
    GPU:
    EVGA 1070Ti Black
    Noctua Make some of the best if not best Air coolers they even put water cooling to same in some cases.
     
  3. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    Hi there

    I would agree with you,I expected around 2-3C better temps but not such big decrease in temps

    Currently I'm running 4.5GHz at 1.275v and load temps pretty much same as with H100i at 4.4GHz

    [​IMG]

    Hope this helps

    Thanks,Jura
     
  4. tsunami231

    tsunami231 Ancient Guru

    Messages:
    14,750
    Likes Received:
    1,868
    GPU:
    EVGA 1070Ti Black
    AIO water cooling isnt all it cracked up to. This Current build us R5 Define case I have with the NH-D14 and 3 case fan all running under 1k rpm and it is quieter my slim ps3 and it is silent compared to the ps3 fat/ps4 (not jet engine) and runs cooler then my Old system.

    I some times forget the pc is on I could put all the fans on silent mode and And my load temps on cpu would go up by 3c and seeing cpu barely hits 55c under full load that is saying something for air cooling.

    Then again I dont OC this all stock speeds.
     
    Last edited: Sep 5, 2016

  5. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,750
    Likes Received:
    9,641
    GPU:
    4090@H2O
    Nice comparison, would have thought the AIO to perform better in comparison to an air cooler... either that, or the NH-D15 is really good!
     
  6. Barry J

    Barry J Ancient Guru

    Messages:
    2,803
    Likes Received:
    152
    GPU:
    RTX2080 TRIO Super
    I went from H100i in push pull to Noctua NH-DH15 and my temps are much better

    I was worried my small Haf 912+ case would reduce the ability of the Noctua but it works so much better than the H100i and is very quiet my CPU @1.275 4.6ghz stays mostly under 60c gaming
     
  7. nhlkoho

    nhlkoho Guest

    Messages:
    7,754
    Likes Received:
    366
    GPU:
    RTX 2080ti FE
    I did the same. And its quieter too
     
  8. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    Hi there

    I thought so I will lower at least 2-3C under heavy load and I didn't expected such drop under load

    I've tried around 8-12 pair of fans,from Corsair to BeQuiet,Noctua and Noiseblockers and several others,I spent more on fans than on actual cooling,tried several thermal pastes too on H100i and still small decrease of temps

    Most of reviewers test AIO on open bench and not in actual cases and sometimes will come to play,marketing too,they simply want to get their products too for reviews

    Most of people who switched to proper air cooler have experienced same as I do,lower temps and cooling is much quieter under load and idle

    Noctua are in my view best air coolers,but out there are several others which are within 1-2C like Phanteks PH-TC14PE,Cryorig R1 Ultimate and several air coolers from Thermalright too

    I will be swapping fans for Thermalright TY-147 fans as I don't like the characteristic of Noctua fans

    Hope this helps

    Thanks,Jura

    I tried on my H100i several fans,push or pull and both and difference has been abysmal or so marginal,I decided to get to try Noctua NH-D15 and I will have too for testing R1 Ultimate too with which I want to compare with D15

    But yes agree is quieter,but this depends on fan profile and CPU too,on my I'm running the fans at 1100-1200RPM and on those RPM fans are louder,due this I went with Thermalright TY-147 fans as next upgrade

    Hope this helps

    Thanks,Jura

    I think more people over here will realize its not worth to have AIO on their PC as with good air cooler PC will be quieter and will cool as AIO,although with some coolers you need to have bit more space,but still its worth it


    On X99 and my ASRock X99 Extreme6 compatibility is issue,I can't use GPU in first slot and my GPU is running at PCI_E 3.0 x8:wanker:

    Hope this helps

    Thanks,Jura
     
  9. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,750
    Likes Received:
    9,641
    GPU:
    4090@H2O
    Well I personally wondered if I was wrong to choose a custom loop over AIO, but my NH-D15 is still in use, cooling the power hungry i7-950 in my HTPC :D

    But yeah, I guess AIO's aren't really worth it then... kind of underwhelmed by their relative performance (and possible leaks).
     
  10. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    You are not been wrong to choose custom loop over AIO,I wanted to do WC loop,which I still want to do as right now loudest thing in my PC is GPU which is cooled by AIO and pump noise is killing me ..

    At least you are reused yours D15 and i7-950 is power hungry plus this CPU runs bit hot too

    If you have spare money and want to do own test,then I would recommend get one AIO and test that against yours D15 and you will probably come to same conclusion as we're,but I don't recommend to do that,as it not worth swap D15 for AIO

    What temps are getting on yours custom loop under load ? And at what RPM are you running fans ?

    Thanks,Jura
     

  11. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    Done quick dirty OC today and here are results

    4.5GHz on NH-D15 with ambient temp 21C

    [​IMG]

    4.6GHz on NH-D15 ambient temp 21C

    [​IMG]

    4.7GHz I think is possible,but I think I will need go beyond 1.35v on my 5820k

    Cache ratio seems on X99 Extreme6 or on Extreme series is somewhere in region 35-38x,looks like my Extreme6 doesn't have OC socket and due this I'm unable to run higher Cache ratio,bummer

    Thanks,Jura
     
  12. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,750
    Likes Received:
    9,641
    GPU:
    4090@H2O
    Fans run at ~1100rpm, but only 6 of the 8 on my rads, I guess my fan splitter is broken. Too deep into the machine to easily exchange so unfortunately my temps are going for the 60-65°C too (Pump running at ~55% RPM, 1.3v on the CPU). Cooling my CPU and GPUs, GPUs hardly go above 51°C.

    There's still room for improvement though, some things to toy with in regards to the loop... will improve the next time I do maintenance (either next spring or, if a 1080Ti comes out... :D )
     
  13. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    Hi there

    Yours temps are OK,those CPU temps under heavy load or they're when you are game ?

    Regarding the fans,yes I'm running at same RPM like you are running and at least GPU is cooled properly,in my case I'm running EVGA Hybrid and temps depends on game,like is GTA V,my temps are in 55-62C and in rendering temps are in 50's too,but what I hate on EVGA Hybrid is pump noise

    I'm planning to upgrade later this year to 6950x and Titan Pascal and hopefully then I will be happy,but for how long who knows :wanker:

    Here are pics of my NH-D15

    [​IMG]

    [​IMG]


    Thanks,Jura
     
  14. fantaskarsef

    fantaskarsef Ancient Guru

    Messages:
    15,750
    Likes Received:
    9,641
    GPU:
    4090@H2O
    No those temps are bench temps, under gaming it's around 60°C. It's fine for the CPU, and I could get higher clocks with 1.3v I think, maybe even hit 4.6, but I haven't bothered so far. Luckily the D5 I've got has an acceptable noise at my rpm, so the fans completly hide it's hum.
     
  15. zorandesign

    zorandesign Member

    Messages:
    40
    Likes Received:
    6
    GPU:
    MSI NX7600GT
    Quieter than PS3? Wow, is the PS3 so loud?
    Because I see HCG750 in your PC, which is everything but not silent. I had it, even the best sample had very loud fan... Except in case you have changed its fan.
     

  16. SHOCKTRUPPEN

    SHOCKTRUPPEN Banned

    Messages:
    89
    Likes Received:
    0
    GPU:
    GTX 980 Strix
    it's not an equal test sure you skewed the test to make your precious noctua seem better but its not. Cpu usage was only 88 percent for your noctua and 100 percent for the cooler. Along with other anomalies with pcie and memory etc. Not a very well run "equal" test. I'll believe the test results from Guru3d in reference to aio's and air coolers. Not biased fan boys getting the results they want.:thumbdown
     
  17. jura11

    jura11 Guest

    Messages:
    2,640
    Likes Received:
    707
    GPU:
    RTX 3090 NvLink
    I skewed test,check SIV64 result and check SIV64 100% CPU usage in both cases or on both CPU coolers

    This is on Noctua at 4.5GHz

    [​IMG]

    At 4.5GHz my temps on H100i V2 has been in 80's on PKG and that's with fans running at 1450RPM

    PCIE speed,did you ever owned NH-D15 and X99,then you know,you can't have fitted Noctua with GPU running in first slot which is x16 and I'm running 5820k and this CPU is only 28 PCI lane CPU and due this my GPU is running at x8

    Get yours own fact together and stop it,I'm far of from fan boy

    And mainly check other forums and people own results and reviews

    Regarding the any review,check how they do reviews,usually all are tested on benches and not in actual cases,check here test how is done

    Hope this helps

    Thanks,Jura
     

Share This Page