Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Intel unveils Alder Lake, coming November 4

2»

Comments

  • RidelynnRidelynn Member EpicPosts: 7,383
    laserit said:
    You have to kind of think of the airflow like a stream with back eddies. You want to move all the air you can. It doesn't have to be cold air, 80 deg is ok, its a lot cooler than what its moving, it just needs to move, the less the dead pockets the better. The CFM (cubic feet per minute) and that you don't just have a clean stream that goes in one side and out the other is what's important. The best situation is if all the air moves. You can see this with smoke.
    Another part of that is that, at 8000 ft, the air isn't as dense, so fans aren't quite as capable as they would be closer to sea level because the air just isn't there.

    But yeah, even at 80F ambient, you should be able to keep your system cool enough. Some throttling my be unavoidable just due to the totality of the circumstances, but your not that far away from the 72F that is considered the standard.

    Laserit is absolutely right though - so long as you can keep air moving, you shouldn't have any issues. Don't only think inside the case: if your case is stuck under a desk or in a corner, and there's no airflow near the outside of the case, it's much the same problem.
    laserit
  • laseritlaserit Member LegendaryPosts: 7,591
    Ridelynn said:
    laserit said:
    You have to kind of think of the airflow like a stream with back eddies. You want to move all the air you can. It doesn't have to be cold air, 80 deg is ok, its a lot cooler than what its moving, it just needs to move, the less the dead pockets the better. The CFM (cubic feet per minute) and that you don't just have a clean stream that goes in one side and out the other is what's important. The best situation is if all the air moves. You can see this with smoke.
    Another part of that is that, at 8000 ft, the air isn't as dense, so fans aren't quite as capable as they would be closer to sea level because the air just isn't there.

    But yeah, even at 80F ambient, you should be able to keep your system cool enough. Some throttling my be unavoidable just due to the totality of the circumstances, but your not that far away from the 72F that is considered the standard.

    Laserit is absolutely right though - so long as you can keep air moving, you shouldn't have any issues. Don't only think inside the case: if your case is stuck under a desk or in a corner, and there's no airflow near the outside of the case, it's much the same problem.
    I‘ve thought about air pressure with many things, not on computers though, great point on the fans at altitudes.

    "Be water my friend" - Bruce Lee

  • laseritlaserit Member LegendaryPosts: 7,591
    edited November 2021
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m leaning towards something being faulty.

    Could could be something like the closed loop cooler not having enough coolant.

    Have to swap it to make sure.

    Bios and all that type of thing is al up to date. System has 0 over clocking.

    1600 watt power supply adds zero thermals, really nice set up.

    I’ve got some great ideas for a high end case ;) made in Canada

    edit: There is also Value Added, I’ll need circuit boards and other odds and ends, maybe software. Another small business or four can supply me with those things.

    Somebody have a new chip factory under construction on NA soil yet?
    Post edited by laserit on
    [Deleted User]Ridelynn

    "Be water my friend" - Bruce Lee

  • RungarRungar Member RarePosts: 1,132
    laserit said:
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m leaning towards something being faulty.

    Could could be something like the closed loop cooler not having enough coolant.

    Have to swap it to make sure.

    Bios and all that type of thing is al up to date. System has 0 over clocking.

    1600 watt power supply adds zero thermals, really nice set up.

    I’ve got some great ideas for a high end case ;) made in Canada

    i made my own case. 36"x24"x12". Its a beast! 80% of the front face is a stainless steel very fine mesh screen.  2 fans blow directly over and under the gpu and another fan blows directly over the cpu. A fan at the top of the case blows all the hot air out.  
    laseritQuizzical
    .05 of a second to midnight
  • laseritlaserit Member LegendaryPosts: 7,591
    laserit said:
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m lean towards something being faulty.

    I once had the issue of the CPU cooler not sticking properly to the CPU. I added some rubber padding to the screws holding the CPU part of the liquid cooling and everything is ok since then.
    That’s one part of the AMD design that I don’t like. I don’t like the clamp design and that’s just my tool & die experience thinking.
    [Deleted User]

    "Be water my friend" - Bruce Lee

  • QuizzicalQuizzical Member LegendaryPosts: 25,499
    laserit said:
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m leaning towards something being faulty.

    Could could be something like the closed loop cooler not having enough coolant.

    Have to swap it to make sure.

    Bios and all that type of thing is al up to date. System has 0 over clocking.

    1600 watt power supply adds zero thermals, really nice set up.

    I’ve got some great ideas for a high end case ;) made in Canada

    edit: There is also Value Added, I’ll need circuit boards and other odds and ends, maybe software. Another small business or four can supply me with those things.

    Somebody have a new chip factory under construction on NA soil yet?
    That's why I said to check your fans.  A computer that is fine when you build it can overheat quickly if a fan dies.  Alternatively, getting too much dust can cause problems.  If a heatsink is clogged with dust so that air can't flow through it, that makes it far less effective at cooling.

    Maybe you already did that.  You should be able to feel to see about how hot the air coming out of your case is.  But a 350 W video card is going to put out a lot of heat under heavy loads.
  • laseritlaserit Member LegendaryPosts: 7,591
    edited November 2021
    Quizzical said:
    laserit said:
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m leaning towards something being faulty.

    Could could be something like the closed loop cooler not having enough coolant.

    Have to swap it to make sure.

    Bios and all that type of thing is al up to date. System has 0 over clocking.

    1600 watt power supply adds zero thermals, really nice set up.

    I’ve got some great ideas for a high end case ;) made in Canada

    edit: There is also Value Added, I’ll need circuit boards and other odds and ends, maybe software. Another small business or four can supply me with those things.

    Somebody have a new chip factory under construction on NA soil yet?
    That's why I said to check your fans.  A computer that is fine when you build it can overheat quickly if a fan dies.  Alternatively, getting too much dust can cause problems.  If a heatsink is clogged with dust so that air can't flow through it, that makes it far less effective at cooling.

    Maybe you already did that.  You should be able to feel to see about how hot the air coming out of your case is.  But a 350 W video card is going to put out a lot of heat under heavy loads.
    Everything is in tip top working order as far as one can tell. Rpm’s according to the bios seem fine. Airflow is good.

     If I knew what temperature range was normal if I put a thermometer directly on my gpu’s back plate I would buy one suitable for the purpose.

     I’m going to swap cooling loops and take that out of the equation, for piece of mind.

    Its not like I’ve never had to return something because it didn’t work right ;)

    edit: ambient temperatures are no longer suspect.


    "Be water my friend" - Bruce Lee

  • RidelynnRidelynn Member EpicPosts: 7,383
    edited November 2021
    laserit said:
    Quizzical said:
    laserit said:
    A good case. Very important.
    And then: a large input fan on the left side, on extractor on the top, and a "cheap" liquid cooling which will also extract air from the case through its fan, and you're good.
    Even at 40°C in summer I never had overheating problems, and it's not an expensive cooling solution.
    My GPU never passes 65°C in New World. The CPU is just chilling.
    If you see my case and how it’s configured and located. You would be scratching your head on the temperatures. That why I’m leaning towards something being faulty.

    Could could be something like the closed loop cooler not having enough coolant.

    Have to swap it to make sure.

    Bios and all that type of thing is al up to date. System has 0 over clocking.

    1600 watt power supply adds zero thermals, really nice set up.

    I’ve got some great ideas for a high end case ;) made in Canada

    edit: There is also Value Added, I’ll need circuit boards and other odds and ends, maybe software. Another small business or four can supply me with those things.

    Somebody have a new chip factory under construction on NA soil yet?
    That's why I said to check your fans.  A computer that is fine when you build it can overheat quickly if a fan dies.  Alternatively, getting too much dust can cause problems.  If a heatsink is clogged with dust so that air can't flow through it, that makes it far less effective at cooling.

    Maybe you already did that.  You should be able to feel to see about how hot the air coming out of your case is.  But a 350 W video card is going to put out a lot of heat under heavy loads.
    Everything is in tip top working order as far as one can tell. Rpm’s according to the bios seem fine. Airflow is good.

     If I knew what temperature range was normal if I put a thermometer directly on my gpu’s back plate I would buy one suitable for the purpose.

     I’m going to swap cooling loops and take that out of the equation, for piece of mind.

    Its not like I’ve never had to return something because it didn’t work right ;)

    edit: ambient temperatures are no longer suspect.



    If the fan is moving air, the case has adequate ventilation, and the fins are clean, then the only other component in the chain there is the compound/mating surface. It wouldn't be the first time someone left the plastic protective film on the back of a heat sink, or something similar.

    In an AIO water cooler, you also have the pump and liquid, which you alluded to. You should be able to check AIO stats somewhere - most AIOs have monitoring software that let you check the speed of the pump and temp of the liquid. The liquid should be heating up - showing you that the energy is getting transferred into the fluid, and that means the radiator should also be warm to the touch (roughly the same temp as what the liquid is reported as).

    If the coolant is getting really hot near the component but no where else, then it's possible you have a kink or blockage in the line.

    If all the lines and radiator are getting really hot - check the radiator air side, make sure the fins are clean, and that the fans are going and in the right direction.

    It would be pretty hard to tell what is normal from just a backplate - they aren't really meant to do any cooling, they are mostly just there to add some mechanical stability.
    laserit
  • QuizzicalQuizzical Member LegendaryPosts: 25,499
    olepi said:
    Intel basically cheated and overclocked their chip to use over twice as much power as the AMD chip. In that configuration, with the AMD part not overclocked, Intel can win.

    But that is using a 7nm Ryzen 3 chip. AMD is scheduled to release a Ryzen 4 chip using the 5nm process next year. Ryzen 4 is faster than Ryzen 3, and 5nm is faster and uses less power than 7nm.

    So Intel better be ready to release the successor to their new chip, next year.
    It's only cheating if you give performance in one benchmark and power draw in a different benchmark.  That allows the hardware to ramp up performance and power for the former, then scale them back for the latter.  Far too many sites measure power only in some artificial stress test that doesn't also measure performance, and then drivers can figure out to throttle back when that particular stress test is running.

    That's why it's so important to measure performance, power, and temperatures in exactly the same benchmark.  If a Core i9-12900K is 10% faster than a Ryzen 9 5900X, but takes an extra 100 W to get that performance, then it's quite reasonable to prefer either part.  I think that the best solution is to have an adjustable power cap, so that the end user can decide how much power use is acceptable.  Some embedded or laptop parts have configurable power usage like that.
Sign In or Register to comment.