Originally Posted by Belial
Originally Posted by wermad Warning: Spoiler! (Click to show)
I'm sure someone will find fault since its coming from Tom's (
). But hey, at least its recent
. Definitely going with Gelid Extreme
at the "alternatives"
Yes, Tom's hardware is a joke. It's pretty obvious why, look at the results - they are testing at 30C load temps? Gee, how indicative of real world overclocks
It'd also help if they were using a hot
CPU - you know, not some dualcore Athlon
Put it on a real world 80c+ real world overclock, and that 2C difference will be more like 10c.
The GPU test comes close
to being useful, and the resuls are exactly as I stated - CLU has a 7C gain over AS5
. I'd say that's massive for thermal paste, you'd never see a 7C temp drop anywhere else in the cooling world with only $10 bucks difference. That's the best value you'll find in cooling anywhere
. Even a Delta fan won't get you 7C!
We're talking about pastes that cost about $2-5, so yea, it's a huge world of difference for 1-2C temp gains for just a dollar basically. CLU is a league of it's own, and it's definitely a lot more expensive, but as you can see in the GPU test results, The difference between it and #2 Gelid Extreme, is the same gap between #2 and last place TIM.
It's not the worst review I've seen, but testing such a low overclock on low heat makes it not very useful. But exactly as I said, there's little difference between modern, high end ceramiques
like Gelid, PK-3, Masscool. It's just the awful pastes that are far worse, like As5 being 5C worse than Gelid, and the extreme TIMs like CLU where it's 3C better, where you see dramatic differences in thermal pastes. Seeing as how CLU is only $15, it's really the best value you'll find in cooling, to use CLU. You'd never get as much of a temp gain for only $15 as you would with CLU, not $15 extra in a heatsink or water cooling parts, not $15 extra in fans...
Put it on a i7-4770K/3770K, an 8350, even an i5, or god forbid an Intel i7 Extreme, and you'll see the 10c+ gains I am talking about just with switching thermal paste. As usual, tomshardware is a joke.
What's scary though, is that some people think testing at 30C load temps with a dualcore athlon is somehow indicative of performance in a real world overclock