Cooling

Discussion in 'hardware' started by Rico, Mar 30, 2017.

  1. hawki

    hawki Registered Member

    Joined:
    Dec 17, 2008
    Posts:
    6,065
    Location:
    DC Metro Area
    You might want to avoid EVGA GTX 1070s, as many have reported overheating issues with them. EVGA is supplying thermal heat pads to those who request them. (Despite the foregoing,EVGA maintains that under normal operating conditions there is no overheating issue.)

    Are you sure 550 W is sufficient? The minimum recommended power for a single 1070 is 500W. Or are you adding the 550W in addition to your current PSU?
     
    Last edited: Apr 23, 2017
  2. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    I went with two Giagabyte 1070's one rear blow, the other a short card. I do remember hearing about problems with overheating & fire with evga. Also I bumped up PSU to 750W. Using this http://outervision.com/power-supply-calculator gave the 550W, as a hedge went with 750. 750 is sole provider, both cards recognized. The DC project has them max'd out.

    Thanks!
    Go Capitals? Luv the Nats
     
  3. plat1098

    plat1098 Guest

    Another source for odds 'n ends. Very nice to deal with and prompt, even free, shipping. .

    https://www.moddiy.com/
     
  4. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    Thanks plat1098 now bookmarked!

    This old Dell server, with the two GTX 1070's. Is running DC project & doing well. Speccy routinely gives the following temps

    Intake air 21 C, CPU 76 C, Graphics 82, 68 C. Do I have more tinkering to do, or is this Okay

    Note: The cool GTX 1070 is the one with one fan, blowing out the rear.

    Thanks
    Rico
     
  5. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    If that is the peak temperature and most of the time, it sits lower, then no problem. But if the CPU's temp is sitting up there for more than a few seconds at a time, 76°C is a little high for me. It is too close to the maximum Tcase allowed of 80°C. While technically, 76°C is not excessive, and the processor's thermal protection features will prevent permanent damage, prolonged exposure to higher than normal temps can increase aging to the device, and the surrounding components. But "can" does not mean "will". It may still provide many years of service but if me, I would like to see it down a few more degrees.
     
  6. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    Thanks Bill! Do u have any suggestions for a cpu cooler, that world work with that CPU, I would imagine the cooler the better as that would help lower the hot graphics card's temp.

    Also
    The blow hole on the case, is sucking air in, using a 120mm fan, salvaged from the old case. It can accommodate a 140mm fan. I was thinking of making it an Exhaust fan, & any suggestion for a 140mm fan.
     
  7. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Graphics card temp? I specifically said the CPU's temps concerned me, if it was sitting at 76°C for extended periods of time. I did not say anything about the graphics card. GPUs run hotter than CPUs.

    And "the cooler the better" is not true. Adequate cooling is absolutely essential. But, for example, a CPU running at 40°C will not perform better, be more stable, or have a longer life expectancy than the same CPU running at 55°C.
    :eek: :confused: Now think about that for second. A "blow hole" fan is called a "blow hole" fan because it located on the top of the case like the blow hole on a whale. Heat naturally rises. So does it make sense to work against nature and try to force the rising heat back into the case? No!

    You want front-to-back and/or bottom-to-top flow of air through the case. Turn your blow hole fan over so it exhausts heated air out and see how your temps are.
     
  8. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    Hi Bill, The sentence just before your quote of me. Reads
    The mention of graphics thinking is a lower CPU temps would lower the heat load, thus causing cooler GPU temps.

    76 deg C CPU, for the Dell server Xenon, seeking your expert advise for a CPU cooler for, this constantly maxed out, machine.

    Thanks
    Rico
     
  9. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Your CPU cooler is only throwing heat into the computer cavity so a more efficient CPU cooler will, if anything, throw more heat the GPU cooler will grab. In other words, it will not help your GPU cooling.

    I am a fan (no pun intended) of OEM coolers so I don't have much experience with aftermarket coolers anymore. I like the EVOs. Just use google and look for reviews.
     
  10. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    2017-05-07_7-51-44.jpg
    Note: Where it says "not current air flow" reverse arrow for current air flow direction. This was a test of Delta more powerful fan acting as exhaust.

    Note: Also fan barley visible (bottom left pic) was acting as exhaust at time of pic. Now blowing in as arrow indicates.

    Case orientation: PSU at bottom of case, "not current air flow" is Blowhole CD rom, HDD is front.

    Current temp: intake air 23 C, CPU 76 GPU 82 67

    Surprise! I tried more powerful fan, acting as exhaust and intake at BLOWHOLE, & intake/exhaust, bottom of pic fan. Within 10 minutes CPU was heading north of 90C.

    I then put the old less powerful fan in as INTAKE at the blowhole, with all other fans blowing in the direction of arrows. With case cover removed, very slowly CPU temps dropped. Put the case cover on. After approx. 3 hours, CPU temp went from 90+ to 76C.

    As I write this, all fans intake, psu exhaust: cpu 76 gpu's 82 67C

    Another surprise, I thought the long rear fan, GPU would be the cool one. The short one is consistently 15 de cooler than the other.

    The case has room for two more fans, blanked off by mfg.

    Any thoughts on reducing temps?

    Thanks
    Rico
     
  11. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    What do you mean intake at blow hole? I fear you are ignoring advice. (1) Case "blow holes" are only located on the top of cases. Fans in any other location are not blow hole fans. They are front fans, side fans, rear fans or bottom fans, but not blow hole fans. (2) Blow hole fans always "exhaust" heated air out. They are NEVER intake fans.

    Also, the case interior looks coated in heat-trapping dust. And it may be just the angle but I don't see a fan mounted on the CPU cooler.
     
  12. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    No confusion here! Blowhole = top of case, or where it says in yellow font "not current air flow" FULLY UNDESTOOD TOP=blowhole, at time of previous post.

    Look at the pic where it says "not current air flow" the arrow points to exhaust + more powerful fan, so NO I did not ignore advise.

    As mentioned above using (current fan orientation) all intake, except psu. It took about 3hrs to go from 90+ to 76 C. I think I did not wait long enough with the more powerful fan acting as exhaust!

    QUESTION Should the fan (partially visible, bottom left of pic) also be exhaust as well?

    QUESTION Do you think more powerful fan, for exhaust BLOWHOLE, or less powerful fan? See pic. might negate rear 140mm fan's purpose as exhaust fan

    NOTE: I would like to make the changes, at the same time, hence ( QUESTIONS ) then wait, to see the outcome after several hours

    ALSO Top of pic "FAN" is blowing at CPU, this is how it was in the old case, the new case did not provide fan "Mounting" screw holes, so it was TY-Wrapped, so as to be in same position, as original. That's why (previous post) I asked about CPU cooler.
     
  13. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    :( As I and others in this thread have repeatedly said several times before, you want front-to-back and/or bottom-to-top flow. With that, YOU should easily be able to determine which direction that fan and every case fan should be blowing - either to facilitate cool air intake, or exhausting heated air out.

    Fans will not negate each other if they support that front-to-back and/or bottom-to-top flow.

    I think you have enough information to take care of this but I will say two more things. (1) My original recommendation from post #2 still stands and that is you would be better off buying a new case that provide proper and sufficient cooling options and (2), cases channel air "flow". When you remove the case cover/side panel, that "flow" through the case is disrupted. It is like blowing into a tube versus blowing into an open room. So typically, you get better component cooling with all case covers/panels properly attached.
     
  14. Rico

    Rico Registered Member

    Joined:
    Aug 19, 2004
    Posts:
    2,286
    Location:
    Canada
    I have bottom to top flow, what to do with rear fan, a) turned off or b) also exhaust or c) intake.

    This a NEW Fractal Design Define case. From before post #48 discussion, online in new case @post #48
     
  15. Hadron

    Hadron Registered Member

    Joined:
    Apr 1, 2014
    Posts:
    2,137
    It's been awhile since I built a new computer. A lot of the motherboards now support water cooling.
    Is it necessary to use water cooling for the latest Core i9 CPUs?

    In the past I have just stuck with quality quiet air cooling such as Noctua.
     
  16. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    No. But this question really should have been posted in a new thread instead of dredging up a thread that has been dormant for over 3 years.
     
  17. Hadron

    Hadron Registered Member

    Joined:
    Apr 1, 2014
    Posts:
    2,137
    Thanks, Bill. :thumb:

    Nothing wrong with this thread.
    It's still working perfectly fine.
     
  18. Hadron

    Hadron Registered Member

    Joined:
    Apr 1, 2014
    Posts:
    2,137
  19. longshots

    longshots Registered Member

    Joined:
    Oct 20, 2017
    Posts:
    533
    Location:
    Australia
    Always been my choice and have never seen a good reason to change.
     
  20. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Not the point. The point is, you should have started a new thread instead of (1) dredging up an old thread and more importantly (2) hijacking someone else's thread for your issue. Hijacking another's thread is considered poor forum etiquette.
     
  21. longshots

    longshots Registered Member

    Joined:
    Oct 20, 2017
    Posts:
    533
    Location:
    Australia
    I agree. Totally wrong.
    I looked back to see where it happened, and would you believe it occurred at post number 2 - the first reply to the question.
    Which, by the way, was posted by you.
    Which, by the way, is a very complete and informative piece of information about tower cooling, and a good read.

    And the first reply was by mirimir.
    There was nothing said at the time about hi-jacking then so that set the theme for future posts.

    Look, I'm sorry, but the thread is
    Forums>Software, Hardware and General Services>hardware and the title of the thread is Cooling, and the question relates to the topic.
     
  22. Bill_Bright

    Bill_Bright Registered Member

    Joined:
    Jun 29, 2007
    Posts:
    4,041
    Location:
    Nebraska, USA
    Oh, bullfeathers. (1) My post #2 was made on the same day as the OP's opening post. (2) Rico (the OP) was specifically talking about pressures and CFM, which my #2 post addressed. (3) Hadron came in a full 2 years later asking about his i9 which had nothing to do with the OP's question.

    So Hadrons question should have been made in a new thread.

    You might want to check out item 15 on Twenty Five Forum Posting Etiquette Tips.
     
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.