1. OCAU Merchandise now available! Check out our 20th Anniversary Mugs, Classic Logo Shirts and much more! Discussion here.
    Dismiss Notice

Radiator Experiment - Results only!

Discussion in 'Extreme and Water Cooling' started by mrbean_phillip, Jan 12, 2014.

  1. mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    Hello,
    Please keep this thread for the results only, if you want to comment, please DO SO HERE

    Cheers,
    Beano
     
  2. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    Test 1 -GPU
    Furmark Burn-in

    Alphacool UT60 280mm Radiator
    DDC 3.20 Pump, 4,347 rpm (note pump max operating temp to be at 60 deg C - 140 deg F)
    Flowrate = 3.31 lpm

    Fans 2x Cougar CF-D14HB-R @ rated speed

    Ambient air @ 23.1 deg C

    GPU Tests
    Rad hot side water -> 48.9 deg C
    Rad cold side water -> 43.7 deg C
    Rad exhaust air -> 44.4 deg C

    GPU 1 -> 58 deg C
    GPU 2 -> 58 deg C
    GPU 3 -> 60 deg C

    CPU Core 1 -> 65 deg C
    CPU Core 2 -> 60 deg C
    CPU Core 3 -> 65 deg C
    CPU Core 4 -> 64 deg C
    CPU Core 5 -> 65 deg C
    CPU Core 6 -> 62 deg C
    *CPU load at an average of 18%

    *Note temps might be varying by 0.1-0.2 Deg at equilibrium....

    This is the most extreme condition you will attain testing GPU burn-in, and even if ambient should be sitting 25-30 deg C, for the purpose of burn-in testing, you should not need more than an hour to ascertain if the GPU is stable or not.

    As you can see, GPU's are very happy and comfy. I am running triple Gainward 780GTX GLH editions, probably the highest factory clocked GPUs you can get.....
    edit1:Re-running this one, ambient's gone up a little, plus I have forgotten to take a screenshot...face-palm....
    edit2: Results updated, screenshots to follow...
    [​IMG]
    and
    [​IMG]
    and
    [​IMG]
    With the Furmark test, I have been running it for around 45 minutes, stopped it for a second to clear the FPS min/max/ave and restarted.....water was at equilibrium, very pleasing results wrt temps as can be seen.
    I hope this will put to bed what a 280MM rad can handle in max stressed environment - next we will do some real-life gaming and I suspect drastically lower temps will be the norm...
    System returns back to normal very quickly after tests are aborted, X-axis probably around 10 minutes.....
    [​IMG]
    and
    [​IMG]
     
    Last edited: Jan 12, 2014
  3. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    Test 2 - CPU:
    Prime95 64-bit CPU test

    Running Prime 95 64-bit until equilibrium...this is for CPU testing, a real-life application, loading the CPU to levels you might see in real life, if you are prone to benching....for gaming applications, you will not see CPU reach these levels with current applications..

    Alphacool UT60 280mm Radiator
    DDC 3.20 Pump, 4,360 rpm (note pump max operating temp to be at 60 deg C - 140 deg F)
    Flowrate = 3.3lpm

    Fans 2x Cougar CF-D14HB-R @ rated speed

    Ambient air @ 21.3 deg C
    CPU at 1.224 Vcore, 4.02GHZ....even overclocking this to 4.7Ghz and raising Vcore to 1.40 Volts, will only add around 10 deg C to the equation.

    *Note the Windows Task Manager not picking up speed proper, CPU-Z reads as what is set in the Bios. Also, temps might be varying by 0.1-0.2 Deg at equilibrium....
    CPU Tests
    Rad hot side water -> 34.0 deg C
    Rad cold side water -> 32.0 deg C
    Rad exhaust air -> 29.5 deg C

    GPU 1 -> 35 deg C
    GPU 2 -> 36 deg C
    GPU 3 -> 36 deg C

    CPU Core 1 -> 57 deg C
    CPU Core 2 -> 55 deg C
    CPU Core 3 -> 55 deg C
    CPU Core 4 -> 57 deg C
    CPU Core 5 -> 56 deg C
    CPU Core 6 -> 59 deg C
    *CPU load at an average of 100%

    This is close to the most the most extreme condition you will attain testing CPU burn-in, again noting ambient could go up significantly before water temp might become an issue.

    For the guys running OCCT or Intel Burn-in, please do not run these programs for more than 10-15 minutes. They subject your CPU's to conditions they were not designed for, will never see in real life, and can cause sudden-death syndrome.

    There are programs such as Prime95, Super-PI et al, or even SETI and others which will do a very good job of determining system stability.

    Above tests with the GPU's in idle....
    *One thing to note is how significantly the efficiency of the rad improves once we start getting a decent (read big enough) delta between water and air....in the GPU tests, there is nearly a 5 deg C drop from hot to cold side temps....
    In the CPU tests, with a much lower heat load, there is only a 2 deg C difference....
    Screenshots for CPU testing, seems the resolution allowed by Photobucket is pretty crappy...
    [​IMG]
    and
    [​IMG]
     
    Last edited: Jan 12, 2014
  4. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    Test 3 - GAMING:
    iRacing 5760x1080 all high details

    Been running a practice session with around 10-13 other guys on the track. All graphics settings to highest level, triple Asus 27" 144Hz displays, 5760x1080 resolution.

    In this mode, hardly an impact on the cooling capacity of the 280mm Alphacool, but then again, I always expected the outcome. Shortly I will download and install Futuremark Benchmark, see how that goes when looping. Should be pretty demanding I would imagine.

    Alphacool UT60 280mm Radiator
    DDC 3.20 Pump, 4,360 rpm (note pump max operating temp to be at 60 deg C - 140 deg F)
    Flowrate = 3.3lpm

    Fans 2x Cougar CF-D14HB-R @ rated speed

    Ambient air @ 24.6 deg C
    CPU at 1.224 Vcore, 4.02GHZ....even overclocking this to 4.7Ghz and raising Vcore to 1.40 Volts, will only add around 10 deg C to the equation.

    *Note the Windows Task Manager not picking up speed proper, CPU-Z reads as what is set in the Bios. Also, temps might be varying by 0.1-0.2 Deg at equilibrium....

    CPU Tests
    Rad hot side water -> 35.9 deg C
    Rad cold side water -> 33.7 deg C
    Rad exhaust air -> 34.1 deg C

    GPU 1 -> 39 deg C
    GPU 2 -> 39 deg C
    GPU 3 -> 40 deg C

    CPU Core 1 -> 49 deg C
    CPU Core 2 -> 50 deg C
    CPU Core 3 -> 51 deg C
    CPU Core 4 -> 54 deg C
    CPU Core 5 -> 48 deg C
    CPU Core 6 -> 52 deg C
    *CPU at an average load of 24%


    [​IMG]

    and


    [​IMG]
     
  5. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    Placeholder
     
  6. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    3DMark 2013 installed and licensed, Pro version....

    A couple quick benchmarks and results, I will do a loop test shortly...

    [​IMG]

    and

    [​IMG]

    and another quick run, nothing optimised, lots of stuff running in the background, top 14% of all PC's on Futuremark....blah, and only playing iRacing with it :lol:

    [​IMG]
     
  7. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    And a final 3dMark benchmark......

    I have to say the system is coping with the heatload very well, I had 3dMark looping earlier, but it could not generate close to the same amount of heat as FURMARK.

    Hot and cold water temps were over 10 deg C cooler, and ambient was nearly 27, so I would take FURMARK runs of earlier as the most exhaustive test you would encounter in any gaming environment...

    Below the latest results, there's also a graph of temps, nice and cosy.

    [​IMG]



    [​IMG]




    [​IMG]
     
  8. OP
    OP
    mrbean_phillip

    mrbean_phillip Member

    Joined:
    Jun 18, 2002
    Messages:
    4,917
    Location:
    Doreen (3754) Vic
    @ Menthu_Rae:
    Final tests for today, with flow at a certain 3.28-3.29 lpm. If you can run your calcs again, would be great :) Cheers!

    I am always amazed, that once water hits equilibrium, at how stable all the temps are - I have been running FURMARK for just over an hour here, and it is quite evident, as long as ambient, and the system load, remains stable, the system will remain in balance....beautiful.....falling in love with watercooling all over again.

    Test below should clear it up that this radiator would be capable of cooling any gaming system, equal to or lower in spec, than this one. Very happy with the result, to be honest. All temperatures remain well in spec, even the cold water, being fed to the pump inlet, 26% below rated max for the pump.

    GPUs and CPU all very happy and far below nominal aircooled temperatures....

    Test 3 -GPU
    Furmark Burn-in, 1 hour

    Alphacool UT60 280mm Radiator
    DDC 3.20 Pump, 3,560 rpm (note pump max operating temp to be at 60 deg C - 140 deg F)
    Flowrate = 3.285 lpm

    Fans 2x Cougar CF-D14HB-R @ rated speed

    Ambient air @ 26.32 deg C

    GPU Tests
    Rad hot side water -> 53.49 deg C
    Rad cold side water -> 47.49 deg C
    Rad exhaust air -> 48.31 deg C

    GPU 1 -> 62 deg C
    GPU 2 -> 62 deg C
    GPU 3 -> 65 deg C

    CPU Core 1 -> 67 deg C
    CPU Core 2 -> 64 deg C
    CPU Core 3 -> 64 deg C
    CPU Core 4 -> 70 deg C
    CPU Core 5 -> 64 deg C
    CPU Core 6 -> 66 deg C
    *CPU load at an average of 27%

    *Note temps might be varying by 0.05-0.1 Deg at equilibrium....

    [​IMG]

    and

    [​IMG]

    and

    [​IMG]
     
    Last edited: Jan 12, 2014

Share This Page

Advertisement: