I blame the push pin mounting. Push pins such enough as it is, and from the looks of it, they put the fan on first, then installed the push pins by wedging a finger between the fan and the pin itself. That is just not good form for mounting a cooler, probably just one of the pins is "locked" in place but not actually in all the way causing uneven pressure and ridiculous temps.
hmm... i got 2 of these cuz they were 15 bucks, and put on on my dual athlon 64 4200, and one on my core quad 6600. the athlon machine saw the cpu temp drop 9 degrees c, while the intel one only dropped about 3 degrees. it could easily be that those plastic tabs on the intel style mounts just dont supply enough positive pressure.
Hmm...I see under test configuration that the ASUS silent square pro cooler was supposed to be included in this review, and yet it wasn't. The list in test configuration also seemed to have more coolers than the benchmarks, so I'm sure there were more that were skipped. Can we possibly see the benchmarks for the rest of these coolers?
yeah, but it says ambient noise was measured at 38 db which should be with all systems off, or with only a passive cpu cooler on the setup. Either way, it's too loud.
From the charts it looks like the reviewer did a bad installation of the heatsink. It's very common for push pin heatsink to have improper installation, especially if people are too lazy to take out the entire motherboard for installation. I know, I did this with the arctic cooling freezer pro 7. You can't just slap the heatsink into position and push the pins in and call it proper. There's not enough clearance between the mobo and mobo tray for you to do that, and you'll just get the pins bent. You have to take the mobo out, make sure there's enough clearance behind the mobo, set the heatsink in place, and make sure the pins are properly threaded through the mounting holes, locked, and secured. You might need 3 hands for the installation if you're one of those clumpsy people.
Even with the poorly finished heatsink surface, the temperature shouldn't but that bad. It's obvious one or more corners of the heatsink is lifted during testing due to improper installation. I suggest the reviewer redo the testing using the proper installation method I mentioned above.
Who knows what went wrong with the testing methodology...but I think we can all conclude that there is no way those numbers are correct. Besides, any heatsink review where the Arctic Cooling Freezer 7 Pro places 2nd place in load cooling temps is automatically suspect.
Load temperatures at stock clock speeds on a dual-core CPU are hardly that demanding. That the Freezer 7 Pro has a higher minimum fan speed could easily account for its placement. Look at the overclocking and scaling charts, and of the tested HSFs the Freezer 7 Pro places in the middle of the pack.
This cooler is identical to Xigmatek HDT-S963 that I have, even the typo on the base label is the same. I have mounted the Xigmatek using the Crossbow kit they sell and Arctic Cooling 5 and my temperatures dropped about 7 degrees C when compared to Intel stock cooler. I am running E8400 at rated speed in a cramped case.
The pushpins are not enough for that cooler, it is actually 70 grams heavier than Intel specifications for the pushpins. The Xigmatek Crossbow will fit on the Kingwin, get the Crossbow and retest, please.
There are plenty of heavier coolers that use the push pins and get better results than this. If the stated 410g is correct, the stock Intel HSF is heavier than this cooler.
I'd guess either this cooler is defective or they fill the heatpipes with an insufficient medium.
They review the identical heatsink sold by Xigmatek. The fans on Kingwin and Xigmatek also have identical specifications. So either Anandtech got a defective unit or the cooler was not mounted properly.
I'd agree it is probably a defective unit, but it is possible the Kingwin component choices for the design are different. Regardless, Newegg has the Xigmatek unit for $25-$10MIR, so not much point to looking at the Kingwin anyway.
I beg do disagree, those heatsinks are identical not only cosmetically, it is the EXACTLY the same cooler, it is just being sold under different brands. No material differences between them.
I cannot remember where but I did read somewhere that Kingwin is the actual original maker of these (Xigmatek designed and engineered the cooler), it would not make any sense for them to use better materials for other brands and shoot themselves in the foot by offering an inferior solution.
Hmmm... Granted, this has only a 92mm fan, but I would have expected better results than this. This looks a lot like the top-rated Xigmatek HDT-S1283 and OCZ Vendetta 2. How was the thermal paste applied? Application of thermal paste on these exposed heatpipe coolers does not follow the same bb-sized, rice-sized, or pea-sized glob in the center of the IHS recommendation that you would use for a normal cooler. As another poster mentioned, the bottom of the heatsink looks pretty rough, and a good lapping might be in order, but I don't think that will make up for these horrible results. I use the Xigmatek mentioned earlier with a 120mm fan and like many others, get superb cooling. (Check out the "Cases and Cooling" forum to see what others say.) Does the 92mm fan make *this* much difference? Or does Kingwin implement the cooler design (or use different materials) in a different way than the Xigmatek or Vendetta?
Thermal paste was applied over the entire surface of the CPU, spread to a thin layer. It shouldn't be an application problem, but I will verify when I remove the cooler and take pictures of the surfaces with an article update if I note any problems.
As I mentioned in the article, I was also surprised at the performance, and checked the mounting pins, airflow, etc. several times during testing. It may simply be that we received a bad sample. If I discover any issues, I'll update in the next couple of days.
I just bought a Cooler Master Hyper Z600 cooler with 12 heatpipes and it's awesome! I have a feeling it's gonna outperform the ThermalRight Ultra 120 Extreme and maybe even the Noctua NH-U12P as it is HUGE!! I hope it will be reviewed here soon.
It is supposed to deflect some airflow down towards the power regulators behind the heatsink. With a cooler that blows air down, the power regulators do get sufficient airflow, but when you use a tower heatsing with a side/front mounted-fan, the MOSFETs do not get enough air and the spoiler is supposed to fix it. Not sure if it works, though, because mine was loose and I ended up taking it off lest it falls out and shorts the motherboard.
As a policy we don't lap coolers we test at AnandTech. Buyers should not be required to do after-purchase grinding and sanding of a cooler to match our test results. Also the curve on many cooler plates (the better ones normally) is curved (not flat) by design and lapping can actually make performance poorer. Our philosophy is to test the cooler as received from the manufacturer as much as possible.
Yes we have lapped a few units and compared performance to the unlapped cooler. We may even comment on those lapped results, but test results for comparison are reported for the cooler as received.
There must be sth wrong with the installation. I have seen many users report much lower temperature which is comparable with TRUE. I suspect it's because of TIM apply method. HDT needs different TIM apply method to take advantage of the HDT concept.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
34 Comments
Back to Article
brian_riendeau - Monday, July 21, 2008 - link
I blame the push pin mounting. Push pins such enough as it is, and from the looks of it, they put the fan on first, then installed the push pins by wedging a finger between the fan and the pin itself. That is just not good form for mounting a cooler, probably just one of the pins is "locked" in place but not actually in all the way causing uneven pressure and ridiculous temps.flopsie - Monday, July 21, 2008 - link
hmm... i got 2 of these cuz they were 15 bucks, and put on on my dual athlon 64 4200, and one on my core quad 6600. the athlon machine saw the cpu temp drop 9 degrees c, while the intel one only dropped about 3 degrees. it could easily be that those plastic tabs on the intel style mounts just dont supply enough positive pressure.Dark Legion - Saturday, July 19, 2008 - link
Hmm...I see under test configuration that the ASUS silent square pro cooler was supposed to be included in this review, and yet it wasn't. The list in test configuration also seemed to have more coolers than the benchmarks, so I'm sure there were more that were skipped. Can we possibly see the benchmarks for the rest of these coolers?flashbacck - Wednesday, July 16, 2008 - link
Are those ambient noise readings correct? Ambient noise of ~38 db is pretty loud.Bieszczad - Wednesday, July 16, 2008 - link
Well, with the reported temperatures so high, the fan needs to spin really fast to remove the heat, hence the noise.CoreTemp misbehaving? Bad diode on the motherboard and/or processor?
Flaky BIOS?
Anyway, those temp and noise numbers stink.
flashbacck - Thursday, July 17, 2008 - link
yeah, but it says ambient noise was measured at 38 db which should be with all systems off, or with only a passive cpu cooler on the setup. Either way, it's too loud.Baked - Wednesday, July 16, 2008 - link
From the charts it looks like the reviewer did a bad installation of the heatsink. It's very common for push pin heatsink to have improper installation, especially if people are too lazy to take out the entire motherboard for installation. I know, I did this with the arctic cooling freezer pro 7. You can't just slap the heatsink into position and push the pins in and call it proper. There's not enough clearance between the mobo and mobo tray for you to do that, and you'll just get the pins bent. You have to take the mobo out, make sure there's enough clearance behind the mobo, set the heatsink in place, and make sure the pins are properly threaded through the mounting holes, locked, and secured. You might need 3 hands for the installation if you're one of those clumpsy people.Even with the poorly finished heatsink surface, the temperature shouldn't but that bad. It's obvious one or more corners of the heatsink is lifted during testing due to improper installation. I suggest the reviewer redo the testing using the proper installation method I mentioned above.
major_major - Wednesday, July 16, 2008 - link
Who knows what went wrong with the testing methodology...but I think we can all conclude that there is no way those numbers are correct. Besides, any heatsink review where the Arctic Cooling Freezer 7 Pro places 2nd place in load cooling temps is automatically suspect.JarredWalton - Wednesday, July 16, 2008 - link
Load temperatures at stock clock speeds on a dual-core CPU are hardly that demanding. That the Freezer 7 Pro has a higher minimum fan speed could easily account for its placement. Look at the overclocking and scaling charts, and of the tested HSFs the Freezer 7 Pro places in the middle of the pack.biassj - Wednesday, July 16, 2008 - link
Don't confuse this one with it's bigger brother, RVT-12025. Which is basically the same Xigmatek HDT-S1283. 92mm vs 120mm.Bieszczad - Wednesday, July 16, 2008 - link
This cooler is identical to Xigmatek HDT-S963 that I have, even the typo on the base label is the same. I have mounted the Xigmatek using the Crossbow kit they sell and Arctic Cooling 5 and my temperatures dropped about 7 degrees C when compared to Intel stock cooler. I am running E8400 at rated speed in a cramped case.The pushpins are not enough for that cooler, it is actually 70 grams heavier than Intel specifications for the pushpins. The Xigmatek Crossbow will fit on the Kingwin, get the Crossbow and retest, please.
Tekkie
strikeback03 - Wednesday, July 16, 2008 - link
There are plenty of heavier coolers that use the push pins and get better results than this. If the stated 410g is correct, the stock Intel HSF is heavier than this cooler.I'd guess either this cooler is defective or they fill the heatpipes with an insufficient medium.
Bieszczad - Wednesday, July 16, 2008 - link
http://www.frostytech.com/articleview.cfm?articleI...">http://www.frostytech.com/articleview.cfm?articleI...They review the identical heatsink sold by Xigmatek. The fans on Kingwin and Xigmatek also have identical specifications. So either Anandtech got a defective unit or the cooler was not mounted properly.
strikeback03 - Wednesday, July 16, 2008 - link
I'd agree it is probably a defective unit, but it is possible the Kingwin component choices for the design are different. Regardless, Newegg has the Xigmatek unit for $25-$10MIR, so not much point to looking at the Kingwin anyway.Bieszczad - Wednesday, July 16, 2008 - link
"Component choices for the design are different"I beg do disagree, those heatsinks are identical not only cosmetically, it is the EXACTLY the same cooler, it is just being sold under different brands. No material differences between them.
strikeback03 - Wednesday, July 16, 2008 - link
And you know this how? Has the actual maker of the unit stated that both are absolutely identical?Bieszczad - Wednesday, July 16, 2008 - link
I cannot remember where but I did read somewhere that Kingwin is the actual original maker of these (Xigmatek designed and engineered the cooler), it would not make any sense for them to use better materials for other brands and shoot themselves in the foot by offering an inferior solution.mikemcc - Wednesday, July 16, 2008 - link
Hmmm... Granted, this has only a 92mm fan, but I would have expected better results than this. This looks a lot like the top-rated Xigmatek HDT-S1283 and OCZ Vendetta 2. How was the thermal paste applied? Application of thermal paste on these exposed heatpipe coolers does not follow the same bb-sized, rice-sized, or pea-sized glob in the center of the IHS recommendation that you would use for a normal cooler. As another poster mentioned, the bottom of the heatsink looks pretty rough, and a good lapping might be in order, but I don't think that will make up for these horrible results. I use the Xigmatek mentioned earlier with a 120mm fan and like many others, get superb cooling. (Check out the "Cases and Cooling" forum to see what others say.) Does the 92mm fan make *this* much difference? Or does Kingwin implement the cooler design (or use different materials) in a different way than the Xigmatek or Vendetta?Matt Campbell - Wednesday, July 16, 2008 - link
Thermal paste was applied over the entire surface of the CPU, spread to a thin layer. It shouldn't be an application problem, but I will verify when I remove the cooler and take pictures of the surfaces with an article update if I note any problems.As I mentioned in the article, I was also surprised at the performance, and checked the mounting pins, airflow, etc. several times during testing. It may simply be that we received a bad sample. If I discover any issues, I'll update in the next couple of days.
Max G - Wednesday, July 16, 2008 - link
I just bought a Cooler Master Hyper Z600 cooler with 12 heatpipes and it's awesome! I have a feeling it's gonna outperform the ThermalRight Ultra 120 Extreme and maybe even the Noctua NH-U12P as it is HUGE!! I hope it will be reviewed here soon.ImmortalZ - Wednesday, July 16, 2008 - link
The new king IS the Thermalright IFX-14. Two Ultra120s in a single package.Noya - Wednesday, July 16, 2008 - link
A retail passive cooler will never beat 120mm tower coolers.Mgz - Wednesday, July 16, 2008 - link
the base of the heatsink/heatpipe needs a good lapping, it was in such a terrible shape :(icingdeath88 - Wednesday, July 16, 2008 - link
What's that about the spoiler? Seriously? What purpose could it possibly serve?Bieszczad - Wednesday, July 16, 2008 - link
It is supposed to deflect some airflow down towards the power regulators behind the heatsink. With a cooler that blows air down, the power regulators do get sufficient airflow, but when you use a tower heatsing with a side/front mounted-fan, the MOSFETs do not get enough air and the spoiler is supposed to fix it. Not sure if it works, though, because mine was loose and I ended up taking it off lest it falls out and shorts the motherboard.zebrax2 - Wednesday, July 16, 2008 - link
those results are terriblemmntech - Wednesday, July 16, 2008 - link
First of all, the cooler isn't properly lapped. It looks like somebody took coarse sandpaper over the bottom of it. Could be a bad installation too.strikeback03 - Wednesday, July 16, 2008 - link
Not exactly the reviewers fault that the bottom of the cooler is the way it is. IIRC they have tested all coolers in as-received condition.Clauzii - Thursday, July 17, 2008 - link
With a 'plate' like that, one could mount a Antinov Turbopropeller - probably wouldn't help..Wesley Fink - Wednesday, July 16, 2008 - link
As a policy we don't lap coolers we test at AnandTech. Buyers should not be required to do after-purchase grinding and sanding of a cooler to match our test results. Also the curve on many cooler plates (the better ones normally) is curved (not flat) by design and lapping can actually make performance poorer. Our philosophy is to test the cooler as received from the manufacturer as much as possible.Yes we have lapped a few units and compared performance to the unlapped cooler. We may even comment on those lapped results, but test results for comparison are reported for the cooler as received.
sukhoi37 - Wednesday, July 16, 2008 - link
There must be sth wrong with the installation. I have seen many users report much lower temperature which is comparable with TRUE. I suspect it's because of TIM apply method. HDT needs different TIM apply method to take advantage of the HDT concept.Bieszczad - Wednesday, July 16, 2008 - link
Could you point me to some instructions how to apply the paste in a HDT cooler setup? The cooler booklet was not clear on this matter. Thank you.Bieszczad - Wednesday, July 16, 2008 - link
Google is you best friend:http://benchmarkreviews.com/index.php?option=com_c...">http://benchmarkreviews.com/index.php?o...&tas...
Good read