-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Air coolers....
- Thread starter Cowamundo
- Start date
- Joined
- Apr 8, 2012
- Location
- Hagaman, N. Y.
http://www.frostytech.com/top5heatsinks.cfm
I use FT as my cpu cooling guide mostly....For a long time.
I use FT as my cpu cooling guide mostly....For a long time.
- Joined
- Nov 1, 2001
- Location
- New Iberia, LA
I personally don't think Frosty tests with enough heatload to really separate out the very best coolers from the bunch. 150 watts of heatload isn't enough for today's hot and power hungry quads and hex core (and pseudo octo-core AMD) processors. Their results are germane if you are running stock to near stock, but not with a good overclock on your system. I've seen a big dropoff in cooling performance by quite a few coolers they rate high when you put a 200+ watt heatload on them.
- Joined
- Jul 29, 2011
- Location
- Buenos Aires, Argentina
Noctua nh-d14/c14, True spirit 140, Silver Arrow, Cm tpc new cooler...
- Joined
- Nov 1, 2001
- Location
- New Iberia, LA
The Noctua NH-C14 doesn't belong there. It can't even handle my Bloomfield at 4.0 with The Beast installed, much less the stock fans. And can't handle 3.8 except with high speed fans. Review coming soon.
- Joined
- Jul 29, 2011
- Location
- Buenos Aires, Argentina
Really?
I think it should be of for Sandy Bridge cpus though...Prove me wrong!
At least it fits in smaller cases.
- Joined
- Nov 1, 2001
- Location
- New Iberia, LA
It does fit in smaller cases. And it's very well made and comes witrh good fans. But a Hyper 212+ handles heat better than the C14. It just doesn't have enough fin surface area. And that's why I say that Frostytech's testing doesn't test with enough heatload.
- Joined
- Jul 30, 2009
I still like my Havik 140 and the Macho.
- Joined
- Nov 10, 2005
Still thrilled with the HR-02 I bought from muddocktor. With a san ace fan at 7V I can't hear it, and real-world load temperatures stay below 60C max.
- Joined
- Apr 8, 2012
- Location
- Hagaman, N. Y.
Getting a max oc out of a Bloomfield is nuts if you run 4.0 Ghz 24/7. They run hot.
150 watt load is the norm, 200 watts isn't. 200 watts for a single benchmark run, yeah. Not a 24/7 oc @ 200 watt load. You sound like a juvenille braggart advocating yourself over Frosty tech, they've been around and have followers for a decade.
- Joined
- May 8, 2004
FWIW - A TRUE with a Panaflo 120x38 Medium Speed kept my Bloomfield i7-930 below 75*C at 4.3GHz at sensible noise levels 24/7 for a few years.
(HT Off)
This isn't exactly Mudd's first rodeo, either
They both have their rationale for their own conclusions (Mudd's being much more stringent IMO)...
- Joined
- Nov 1, 2001
- Location
- New Iberia, LA
I use a 200 watt heatload because it's not hard to get there with a quad (or more) core processor and have it stable at that speed for years on end. Besides testing heatsinks, I also run DC projects with my systems and that means 100% load 24/7/365. And there are quite a few people on these forums that do the same thing and with proper cooling and keeping voltages below Intel and AMD's max vcore guidelines you won't hurt anything. My younger brother still has some old Conore dual cores and a Q6600 quad I ran for years overclocked in the 3.4-3.6 range and they are still going strong. If you keep vcore reasonable and keep the proc cool, you can push your proc for a long, long time without failure. And since I have had such good success with this formula, I look for heatsinks that able to handle that kind of running. As for my Bloomfield at 4.0, that processor hardly needs any vcore adjustment to run there. I actually run it at a higher than needed vcore at 1.305 volts to get some extra heat for testing. If it weren't a test system, it would also be running BOINC 24/7/365 at that overclock, as that proc easily handles the speed without passing Intel's max vcore guidelines.
I'd like to throw 2 things out here. Firstly, your post seems a bit harsh and you should re-read what he said because what he said is 100% true and correct and even in my opinion a compliment to frostytech. Also, check the definition of germane if that tripped you up.
The second thing here is to elaborate on what he said. Many people do buy these massive air cooling solutions with no intention of over clocking or even putting them on a high power processor and Frostytech certainly caters to that idea. To use your own words, they cater to "juvenille" braggarts. That's fine in my books because it's certainly a large part of the "performance computer" market, but it's not hard for me to make an argument that the true test of the design quality of a high performance heat sink is how it performs when it is pushed hard. Whether one person's stock AMD bulldozer runs at 29.1C or 29.4C is much less important than whether another person's AMD bulldozer at 5.0Ghz is running at 50C or 80C. I remember seeing at least one review where an 8-core bulldozer was drawing upwards of 300 watts( I think it was over 350) when it was over clocked and I see 3 heat sinks in frosty techs top 10 that I'd bet money on failing to handle the sort of heat that chip would be producing. Actually, I just looked on frostytech and my heat sink is in about 50th place on their scale of heat sinks. I've done a comparison with another forum user who has their 5th place heatsink(NH-D14) and with our bulldozers drawing about 240w we get nearly identical temperatures. If you are a true high performance user who needs something other than the stock heat sink then their reviews are starting to lack relevance with many of the new high-end processors. If you just want to show your friends your cool looking heat sink that doesn't let your stock processor rise more than 5C above room temperature that's perfectly cool with me because I've done that before and FrostyTech can help you make your choice.
- Joined
- Oct 22, 2010
- Location
- Cincinnati
Pushing the community to think critically and take public results from any website with a grain of salt is not juvenile, it is how progress is made....
I for one would like to know how things work when pushed PAST the "norm" if all of the top 10 perform equally @ 150w but only 1 keeps producing results @ 160w then i wanna know which one that is because its BETTER even if i don't need it now, i can then make an informed decision...
The only thin juvenile in this thread thus far is your post, re think it, re word it, and contribute productively instead of flaming
I personally like my old school true over the nh d14. I can easily hit 200w with my cpu that has had a hard life, and loves voltage. The nh d14 had a hard time holding my cpu at 3.9ghz, where as my true had no problem at 4000-4100. I do not own any 25x120 fans suitable for hsf use, so I was useing 38x120s on the true. As a note, I did try those same big fans on the nh d14, and they only helped by 1-2c, it was not able to hold the heat from my cpu. I have an h100 now, its much better then any regular air cooler that I have ever used, but it can barely hold my cpu at 4200. To run 4200mhz, my cpu requires 1.4625v, 4100 needs 1.4v, and 4000 needs 1.35. So hitting 200w is no problem at all. I can probably get closer to 250-300w at full tilt with an app like ibt.
- Joined
- Sep 29, 2004
Please keep in mind that this is Overclokers Forums. There are many people on here that run just a few notches under their max overclock 24/7 so your post bears no relevance here.
While it is certainly true that there are many less experienced people on here it ordinarily easy to pick who is who by the wording of their post, and your reply is adjusted in accordance with this.
High clocks don't much matter to me as video conversion and gaming is about the most my system gets stressed, and I can count on one hand the number of times I do a DVD rip each year. Having said that, I will buy the best component available to me and Mudds reviews I will trust over any others out there. It's good to try keep any personal comments off here unless they're positive also although I know we all have our moments.
While it is certainly true that there are many less experienced people on here it ordinarily easy to pick who is who by the wording of their post, and your reply is adjusted in accordance with this.
High clocks don't much matter to me as video conversion and gaming is about the most my system gets stressed, and I can count on one hand the number of times I do a DVD rip each year. Having said that, I will buy the best component available to me and Mudds reviews I will trust over any others out there. It's good to try keep any personal comments off here unless they're positive also although I know we all have our moments.
- Joined
- Oct 30, 2009
I am grateful to muddy for pointing out just why the frostytech reports of heatsink performance differed so much from reviews on the rest of the web. Generally one will find broad agreement on the performance of heatsinks. Sure, they may differ by a couple or even a few degrees, but frostytech seems to inhabit another universe. Now I know why.
And here I thought they were just showcasing heatsinks they made a higher profit from. Silly cynical me.
I do note that they tend to have a group of people who rush to their defense and push their reports over all others on the web. I've not seen that with any other site. Mostly folks want to see a number of reviews and never trust just one. To find frostytech fanatics is . . . interesting. Quite the phenomenon.
As for my own test system, it's an i7 860 running at 4GHz on 1.32v. I think there's more happening on the cpu silicon of the 8xx's than the 9xx's because the northbridge is part of the cpu. So it gets very hot. Amps? <shrug> Dunno. Guess I'll have to figure that out. Haven't cared because I've been about comparing fans on the same heatsink, not comparing heatsinks.
That said, The D14 and the Megahalems can cool equally well on my rig, but the Megahalems needs a stronger (louder) fan. It makes sense, when you consider comparative square mm.
Muddy, you know TRUE's. Do they have more plates, closer together than the Mega? Seems that Trues benefit more from stronger airflow that other heatsinks, which tend to level out.
And here I thought they were just showcasing heatsinks they made a higher profit from. Silly cynical me.
I do note that they tend to have a group of people who rush to their defense and push their reports over all others on the web. I've not seen that with any other site. Mostly folks want to see a number of reviews and never trust just one. To find frostytech fanatics is . . . interesting. Quite the phenomenon.
As for my own test system, it's an i7 860 running at 4GHz on 1.32v. I think there's more happening on the cpu silicon of the 8xx's than the 9xx's because the northbridge is part of the cpu. So it gets very hot. Amps? <shrug> Dunno. Guess I'll have to figure that out. Haven't cared because I've been about comparing fans on the same heatsink, not comparing heatsinks.
That said, The D14 and the Megahalems can cool equally well on my rig, but the Megahalems needs a stronger (louder) fan. It makes sense, when you consider comparative square mm.
Muddy, you know TRUE's. Do they have more plates, closer together than the Mega? Seems that Trues benefit more from stronger airflow that other heatsinks, which tend to level out.
- Joined
- Sep 29, 2004
Toms Hardware used to have a following like this back before they became the mainstream site they are today. I'm talking back in 98 through the early 2000's.....
- Joined
- Nov 1, 2001
- Location
- New Iberia, LA
Your 870 is a good processor to test with IMO, since it is a hot running 45nm model and like you said, has the NB integrated into the processor, which makes for more heat. I remember the 875K I used to own was a pretty hot running beast too, although it only ran under water.
As for the TRUE, the original models did have closer fin spacing, which is why they responded so well to increased cfm and static pressure. You can still see this in action with the TRUE Copper, which is based off the original TRUE. They did open up the fan spacing on the Rev. C models and cut back the number of fins though, as well as cutting a "V" shaped hole through the middle of the fin stack. That seemed to allow better cooling with lesser cfm and static pressure, but didn't seem to hurt efficiency either with high cfm and static pressure fans.
And right now I am writing up my review of the NH-C14 and it illustrates well how a heatsink can perform well at a lower heatload yet fall flat on it's face after a certain point.
