Best cooler for overclocked QX6700?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Best cooler for overclocked QX6700?
Running an Arctic Cooling Freezer 7 Pro right now on a QX6700 and it's loud, and it doesn't cool well enough on those rare occasions that I have a program push all four cores.
So I want to replace it with something that cools the CPU better, but isn't crazy loud.
I'm thinking an Ultra 120 Extreme with the 1600 RPM Scythe S-Flex (which can spin down to a quiet RPM, but speed up under pressure).
Any better alternative? The Ninja gets mentioned, but it sounds like it has some issues...
I'd prefer to go with one that doesn't require me to remove the motherboard, but I'll do it if it'll work very well. I
So I want to replace it with something that cools the CPU better, but isn't crazy loud.
I'm thinking an Ultra 120 Extreme with the 1600 RPM Scythe S-Flex (which can spin down to a quiet RPM, but speed up under pressure).
Any better alternative? The Ninja gets mentioned, but it sounds like it has some issues...
I'd prefer to go with one that doesn't require me to remove the motherboard, but I'll do it if it'll work very well. I
I think it will be hard to find a good quiet HSF that doesn't require you to remove the motherboard.
The Ultra 120 Extreme seems to be the best candidate for the job, provided it fits on your motherboard. If possible, duct the CPU fan to a case fan or an air intake for maximum cooling.
Which case and motherboard do you have?
The Ultra 120 Extreme seems to be the best candidate for the job, provided it fits on your motherboard. If possible, duct the CPU fan to a case fan or an air intake for maximum cooling.
Which case and motherboard do you have?
I'm using an Asus P5K3 Deluxe motherboard in an Antec Sonata II case. The P5K3 is kind of similar to the Striker Extreme in that the CPU socket is surrounded by heatpipes.
I'm okay with removing the motherboard. Just don't want to. But I will.
Haven't messed with ducting... There are two fans inside the case, one thats sits in front of the drives, and one at the rear.
I'm okay with removing the motherboard. Just don't want to. But I will.
Haven't messed with ducting... There are two fans inside the case, one thats sits in front of the drives, and one at the rear.
I am using them on a P5K Deluxe, and it fits fine. Ditto P5B Deluxe and P5W-DH deluxe.dennya wrote:I'm using an Asus P5K3 Deluxe motherboard in an Antec Sonata II case. The P5K3 is kind of similar to the Striker Extreme in that the CPU socket is surrounded by heatpipes..
I hear you about removing the motherboard Never fun, but worth it in the end.
Thanks, DaveXL and Atmosper!
Put the Ultra 120 Extreme in tonight and it knocked about 20C off my temps compared to the Freezer 7 Pro! And with the Scythe SFF21F fan, even at 1600 RPM it's whisper-quiet.
Overclocked to 3.33 GHz on a 333MHz bus, with 3 hard drives and an 8800GTS in a pretty cramped Sonata case, I'm getting idles in the low 40s and a peak temp of 66C with all four cores doing sustained 100% during video rendering. Not bad at all!
Installation was a snap, other than removing the motherboard.
Thanks again for the advice. This heat sink rocks.
Put the Ultra 120 Extreme in tonight and it knocked about 20C off my temps compared to the Freezer 7 Pro! And with the Scythe SFF21F fan, even at 1600 RPM it's whisper-quiet.
Overclocked to 3.33 GHz on a 333MHz bus, with 3 hard drives and an 8800GTS in a pretty cramped Sonata case, I'm getting idles in the low 40s and a peak temp of 66C with all four cores doing sustained 100% during video rendering. Not bad at all!
Installation was a snap, other than removing the motherboard.
Thanks again for the advice. This heat sink rocks.
There have been some quality control issues with the Ultra 120 Extreme. A fair portion of the units had bases that were not flat, resulting in very poor temperatures:
http://forums.extremeoverclocking.com/t259086.html
Based on Newegg reviews, I assume at least 10% had serious problems such as one side being 0.5mm shorter than the other.
So the model seems to be a bit of a gamble. From the feedbacks I would say the Tuniq Tower 120 may be a safer choice
http://forums.extremeoverclocking.com/t259086.html
Based on Newegg reviews, I assume at least 10% had serious problems such as one side being 0.5mm shorter than the other.
So the model seems to be a bit of a gamble. From the feedbacks I would say the Tuniq Tower 120 may be a safer choice
No issues here with mine, and the results were fantastic.
You can't use online posts as a gauge of the percentage of failures... A huge percentage of people who have problems will post to complain about them. A tiny percentage of people who don't have problems will post to say so.
You can't use online posts as a gauge of the percentage of failures... A huge percentage of people who have problems will post to complain about them. A tiny percentage of people who don't have problems will post to say so.
Last edited by dennya on Wed May 30, 2007 8:41 am, edited 1 time in total.
The idea is not random online posts, but average review scores, which includes both people giving perfect reviews and those giving the poorest scores. Aggregate scoring tends to force widespread patterns to emerge. The more reviews averaged, the more likely it is to represent reality. We are not doubting that you and the majority of users have no issues, but there is a statistical difference between 1% of samples receiving bad reviews versus 10-20%. The issues with Thermalright's manufacturing process for the older Ultra 120 version has been reported previously on SPCR and elsewhere as well.dennya wrote:No issues here with mine, and the results were fantastic.
You can't use online posts as a gauge of the percentage of failures... A huge percentage of people who have problems will post to complain about them. A tiny percentage of people who don't have problems will post to say so.
I will add that one will seldom read about such manufacturing issues on official reviews by commercial websites because the samples supplied were likely hand picked by the manufacturer to eliminate potential problems. To be more accurate, you have to sample a fair number of real life units.
Quality control will play into consideration of computer system builders. Customers in many industries will avoid products based on reliability reports, even if the working samples are potentially better. Doing RMA takes time and can often be a frustrating experience.dennya wrote:I also think it's silly to discount an excellent product because of some initial bad units. Just be sure to buy from a place that has a good return policy in case you get a dud.
To quote from a thread on Anandtech:
"all you people who are complaining that your Ultra-120 E sux wahtever....should've gotten a simple to install, no brainer CPU cooler and forgotten all about this cooling"
"the ThermalWrong Ultra 120 Extreme, I've read about the same quality control issues that you describe over-and-over again, here and elsewhere! Unfortunately, ignorance is bliss. Most ppl believe this HSF is 'the one' -=- the one that's going to clean Tuniq's clock -=- and users are openly attacked if they report ANY problems on these web boards."
"ThermalRight has a good reputation for killer products, but it does seem that this product is very hit-or-miss now."
I think the Ultra 120 Extreme would be a great choice when the buyer would not be upset if their heatsink came with issues that required lapping or returns.
No one can claim that getting an Ultra 120 Extreme will necessarily result in poor temperatures, but the probability of defects will matter to some. Over 1/6 of users on Newegg have rated the heatsink 2 out of 5 or less, a non-trivial portion, in conjunction with other reports. To be fair, the recent inflow of new units may have improved quality control over earlier ones. Personally I prefer the Sunbeam because the company explicitly states that every unit has been tested before shipping, and there have been less reports of issues. Also, RMA of a defective Ultra 120 can take months, as reported here on SPCR: viewtopic.php?t=33819
In any case, caveat emptor.
In any case, caveat emptor.