Question Want to improve my CPU cooling, so what is more effective?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
I currently own a Ryzen 5800x that is cooled by a Noctua U12A. My case is a Meshify C that I replaced the stock fans with Cooler Master Silent 120mm fans and added one more 120mm intake fan, so not I have 2 120mm intake fans in the front and 1 rear 120mm exhaust fan and my CPU now runs a little cooler, at a fixed all-core 4.6GHz at 1.25V my 5800x is now hitting 80C in Cinebench R20, whereas before I replaced the stock fans + added one more intake it was hitting 85C at the same CPU settings. Also at CPU stock settings, I seem to be getting slightly better boosts as my Cinebench multicore score slightly improved, but it's still within the margin or error range, since it's like a 1% improvement, but most of the time that 1% improvement is consistent each time I"m running Cinebench R20. However, I'm still not satisfied with my CPU temperatures at stock settings as it hits 90C within 25 seconds of running Cinebench R20 multicore whereas before the case fan replacement + additional intake fan it was hitting 90C within 10 seconds. I'm looking for even better improved CPU cooling and I was thinking about replacing the 2 Cooler Master 120mm intake fans with 2 Arctic 140mm PWM fans and replacing the rear Cooler Master 120mm exhaust fan with an Arctic 120mm PWM fan. I was also consider just replacing my case with one that has vents in the front or front-sides. I also considered replacing my Noctua U12A with a Cooler Master 240mm AIO (don't want an AIO that needs software installed to run) and top mounting it but I read reviews that showed that the Noctua U12A has about the same cooling performance of some 240mm AIO's and the only way I can install a 280mm or larger AIO is if I install it in the front but I heard that doing it this way causes more heat to come into the case that top mounting it as exhaust because of the radiatior emitting hot air into the case if it's mounted in the front as intake. Am I better off with a top mounted 240mm AIO as exhaust than a front mounted 280mm AIO as intake? So what is best for improving CPU cooling to what I have now?
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
14,323
4,904
136
I have a 5800X and set its PPT to 105W for negligible MT performance loss and zero ST performance loss. It also runs almost 15C cooler.
 

bononos

Diamond Member
Aug 21, 2011
3,883
142
106
I have a 5800X and set its PPT to 105W for negligible MT performance loss and zero ST performance loss. It also runs almost 15C cooler.
Why didn't AMD use a lower PPT as the default, were they looking for every ounce of performance because of competition from new Intel cpus?
 

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,619
136
Why didn't AMD use a lower PPT as the default, were they looking for every ounce of performance because of competition from new Intel cpus?
With a high PPT as the default they cover all bases without extra SKUs. Want lower PPT? The user then can use Eco mode or set PPT to something lower without voiding any warranty.
 

blckgrffn

Diamond Member
May 1, 2003
9,110
3,028
136
www.teamjuchems.com
With a high PPT as the default they cover all bases without extra SKUs. Want lower PPT? The user then can use Eco mode or set PPT to something lower without voiding any warranty.

I enabled eco mode on both of my 5800x - and still they stubbornly hot spot and spin up fans. I probably got sub-optimal silicon. But yes, it is an option and I appreciate its inclusion.
 

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,619
136
I enabled eco mode on both of my 5800x - and still they stubbornly hot spot and spin up fans. I probably got sub-optimal silicon. But yes, it is an option and I appreciate its inclusion.
Eco mode only drops a 5800X to 5600X's level (which also has hot spots) so you could easily go further down (though hot spots will always be there to some degree unless you avoid boosting altogether).
 
  • Like
Reactions: blckgrffn

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
Well, about 3 weeks ago I replaced my 2 CoolerMaster 120 Silent case fans with 2 Arctic P14 PWM case fans for the intakes and the CoolerMaster 120 Silent case fan with an Arctic P12 PWM as a rear exhaust. CPU cooling did not seem to improve at all. Actually, my Cinebench scores are consistently slightly lower after than fan replacements due to the CPU throttling a little more to try to not go over 90C. I was expecting the opposite to happen since these are supposedly better quality fans than the ones I replaced and going from 2x120mm intakes running at 1200 rpm (Q-Fan control off) to 2x140 intakes set to run at 1000 rpm I expected better airflow to the CPU. In fact my CPU seems to be only slightly cooler than when I used the stock case fans (1x120 intake and 1x120 rear exhaust). Also I tested the fans at max speed settings and it did not seem to make any change to the CPU temperature, except for a much noisier system.
 

Dave3000

Golden Member
Jan 10, 2011
1,343
91
91
3 days ago I switched back to my CoolerMaster 120mm fans and now I'm getting about 100 points higher in Cinebench 20 CPU multicore test consistently once again and my CPU does not thermal throttle as much now. I would usually get between 5800-5900 (on a few occasions dropping as low as the upper 5700's) in CB20 with the Arctics and with these CoolerMasters I'm getting between 5900-6000 and never dropping below 5900. This seems strange to me as these Arctic fans are advertised to produce better airflow and static pressure than these CoolerMasters, plus when I had the Arctics installed, I had 2x140mm for intakes (and set them to run at the same rpm's as the 120mm CoolerMasters which is around 1200 rpm). I was thinking that the front intake cutouts of my case were probably designed for 3x120mm fans in mind, rather than 2x140mm fans for intakes, as there are metal strips that block some airflow from the 140mm fans and the fan intake cutouts are sized for 120mm fans and much smaller than 140mm fans, despite this case supporting 140mm fans as the case has screw holes to accommodate 140mm intakes as well. Could that be why I'm getting better CPU cooling with these CoolerMasters (1x120mm exhaust and 2x120mm intakes) vs Arctics (1x120mm exhaust and 2x140mm intakes)? Also when I had my Arctics installed I set them to a fixed 1200 rpm, but then again even setting them to variable speed (based on CPU temperature) did not seem to do anything for CPU cooling, and just made this PC louder. Should I have gone with 3x120mm Arctics as intakes instead of 2x140mm Arctics as intakes due to my cases intake fan cutouts size?
 
Last edited: