TechPowerUp CPU Cooler Test System Update for 2024

continuum

Ars Legatus Legionis
94,897
Moderator

Some methodology changes to be kind of like a few other sites (Tom’s Hardware comes to mind),

But methodology changes or otherwise (which you may or may not agree with- I most certainly have mixed feelings), I do find TPU heatsink and AIO testing particularly useful because unlike many others, they test with not only a modern Intel processor (Raptor Lake, i7-13700K) but also a modern dual-chiplet AMD processor (Zen 4, Ryzen 7900X).
 

Nevarre

Ars Legatus Legionis
24,110
There's a lot of room to nitpick this review, but they did a noise normalization against the Dark Rock Slim, but then didn't test it or any other BeQuiet! products. (The Slim's not a powerhouse cooler but considering how badly the Noctua DH12S did...)

Not including the 7800X3D as the likely price/performance gamer pick and likewise arbitrarily picking the 13700K on the Intel side leaves out a lot of useful information. Including some more midrange CPUs would be really helpful as they included a fair number of budget coolers (say 14600 on the Intel side.)
 
Uses Arctic thermal paste, but not an Arctic cooler as part of their testing...mmm. And yet they throw in the venerable, but overmatched Noctua U12S. I remember when I thought it was a beast on my quad core CPUs.

I find their graphs confusing, as I can't easily tell fan vs liquid cooling apart. I'm assuming that AIO's perform better for the higher end multicore beasts.
 

ikjadoon

Ars Scholae Palatinae
1,371
Initially, I thought 45 dBA was a very high noise normalized choice, until I realized their sound meter is a mere 6" away from the CPU cooler.

But 6" is so close: might it introduce directionality? Say a fan has a wobbly corner and the sound meter is close to that one corner. It might exaggerate the difference that wouldn't be as impactful if the sound meter were 12" or 24" away? I don't know enough about sound measurements to know if that's a problem, though.