OCPulse Heatsink Showdown! Comprehensive Air Cooling Roundup

[lowfat]

very nice review!

[Spawne32]

Hmmmm interesting...thought the HR-01 Plus would do better (probably neck-to-neck with the Ultra-120 Extreme)

Just a noob question, should the Ultra ChillTEC be included? Considering it is a TEC cooler?

The TEC cooler is in there because its part of the air cooling catagory

Spawne

Excellent review. Wish this was available before I purchased 2 weeks ago...

Did you try to measure/account for, mount variability?

I assume this was based on 1 mount, as I didn't see it mentioned.

No unfortunatly i could not do it based on one mount, certain mounts just dont work with certain heatsinks, as far as it being a variable.... not really, in my opinion, unless the mounting system was so bad that it adversely effected performance, it should not have been an issue.

During the second set of tests that im currently gathering, im also gona do price charts, and mounting system info, so if you do happen to select a heatsink based off the list, you know what your getting ahead of time.

[boatasiaus]

Spawne.

My bad, wasn't clear....

I meant, how many times did you remove/replace TIM & each heat sink to the CPU and record results?

Personally, I think you need to remount the TIM & HSF at least 3x, record results & then average....

[Spawne32]

Spawne.

My bad, wasn't clear....

I meant, how many times did you remove/replace TIM & each heat sink to the CPU and record results?

Personally, I think you need to remount the TIM & HSF at least 3x, record results & then average....

3-4, depending on whether i was confident on the results or not ;-) Everytime i removed it, i would check to see if the contact area was ok.

[PhilDoc]

Well all the experts i talked to agreed that the open air would provide the best results for variables like cases, these are computers, not wind tunnels if it was a wind tunnel then yes the concept would apply but in this case, if your case airflow is bad, you can expect higher temps, if your case airflow is good, you can expect temps similar to the results shown. Thats how that works basically.

I agree, wasn't agreeing with the post, just trying to do my best to clarify it, lol. The way I see it a level playing field and the way you did the review is the way it should be done.

[boatasiaus]

3-4, depending on whether i was confident on the results or not ;-) Everytime i removed it, i would check to see if the contact area was ok.

Very nicely done. Would add that to the methodology section

[Spawne32]

All of this stuff will be added in the next update im making notes as we go along for all the information i gotta add

[troisanh]

nice read.

[Spawne32]

nice read.

Im not much on words the main thing i was concerned about was the charts and results

[Spawne32]

Good news guys, the mail just came and dropped off one of my international heatsinks :p this might be new for us USA guys, as they deal exclusively for europe atm.




trying to help bump the knowledge of their products here in the USA

[leo_bsb]

great job.
Did you use just one fan for the Ultra 120x ?

[Spawne32]

great job.
Did you use just one fan for the Ultra 120x ?

Yes, im gona need to elaborate more on what those #'s and such are after the heatsink names are

the First thing you will see after the heatsink name is the fan model the second thing is in what orientation the fan was, if there is more then one fan it will be noted, such as

Thermalright Ultra 120 Extreme D12SM-12 PUSH (1 fan, pushing air through the heatsink)
Thermalright Ultra 120 Extreme D12SM-12 PUSH/PULL (2 fans, 1 fan pushing through the heatsink, the other on the opposite side pulling air through.)

[ryboto]

I was surprised the HR-01 Plus was so far behind, but then again, fin design and mounting system do play a big part in performance.

The HR-01 Plus has fewer fins, less surface area for convection. It's ideally suited for low airflow or semi-passive installations. The fin spacing helps with thermal boundary layer issues when there's less airflow.

[Spawne32]

The HR-01 Plus has fewer fins, less surface area for convection. It's ideally suited for low airflow or semi-passive installations. The fin spacing helps with thermal boundary layer issues when there's less airflow.

I did the testing on the two tonight for passive, both failed miserably, basically they just heated up, and continued to heat up, and even after it passed the thermal limit of the tests, it only cooled down 5c, this is even with a fan simulating an intake fan, blowing from a few inches away across the heatsink.

Im gona try with some low speed yate loons now, see how the numbers turn out.

[ryboto]

I did the testing on the two tonight for passive, both failed miserably, basically they just heated up, and continued to heat up, and even after it passed the thermal limit of the tests, it only cooled down 5c, this is even with a fan simulating an intake fan, blowing from a few inches away across the heatsink.

Im gona try with some low speed yate loons now, see how the numbers turn out.

that sucks. Weird, my Ultra 120(4pipes) could passively cool my 89W Opteron with a 200mhz OC(2.2ghz). Temps never reached more than 55C. I've since switched to an AM2 chip, and until last week, was using the Ultra 120 passive, but I've switched to the HR-01 Plus. It cools the X2 4000+ perfectly fine at stock settings passively, but I'm on an undervolting kick.

edit: strange temps you're seeing. My friend has an Ultra 120 Extreme installed on his B3 Q6600, and when he SMP folds the chip only reaches 44C. He's got one fan mounted on the U120X in a pull position, and the rear case fan acts as an intake feeding fresh air from ~ 4inches away at 900rpm to the other side of the heatsink.

[Spawne32]

that sucks. Weird, my Ultra 120(4pipes) could passively cool my 89W Opteron with a 200mhz OC(2.2ghz). Temps never reached more than 55C. I've since switched to an AM2 chip, and until last week, was using the Ultra 120 passive, but I've switched to the HR-01 Plus. It cools the X2 4000+ perfectly fine at stock settings passively, but I'm on an undervolting kick.

edit: strange temps you're seeing. My friend has an Ultra 120 Extreme installed on his B3 Q6600, and when he SMP folds the chip only reaches 44C. He's got one fan mounted on the U120X in a pull position, and the rear case fan acts as an intake feeding fresh air from ~ 4inches away at 900rpm to the other side of the heatsink.

I can tell you whats going on, first of all you cant compare an opteron to a B3 Quad core, these things produce tremendous heat, AMD's run naturally cool, so thats only expected. As far as your friends quad core, im gona go out on a limb here and say that when he SMP Folds, hes not using all 4 cores. The program is probably not designed to utilize all those cores, probably only two, when he runs the program, have him check and see what the CPU stress is on each core, id be interested to see that. On all my tests im priming all 4 cores using Prime95, so its maximum heat load.

[hardnrg]

please consider adding overclocked and overvolted results to your charts...

I don't really care about stock speed/volts... I want to see how the heatsinks handle under pressure... that TEC sink isn't gonna be the best performer when you add vCore

SMP folding (Linux SMP client) is designed for 4 cores... not sure about the Windows SMP client because it is unstable and inefficient compared to the Linux version so I've never bothered with it

[emoners]

The TEC cooler is in there because its part of the air cooling catagory

Oh ok...thanks Spawne32.

[saapas]

Spawne, isn't that Coolink heatsink exactly the same as in Noctua, the only difference being the fan? Oh, and less heatpipes too.

[ryboto]

As far as your friends quad core, im gona go out on a limb here and say that when he SMP Folds, hes not using all 4 cores. The program is probably not designed to utilize all those cores, probably only two, when he runs the program, have him check and see what the CPU stress is on each core, id be interested to see that.

In task manager, it shows that all 4 cores are being stressed...SMP F@H uses 4 folding cores, and if you have 4 physical cores, each core will be fully utilized. I was there when we set it up, so I know it's fully loaded. The frame times of ~11-15 minutes depending on WU also suggest to me that it's fully loaded. He doesn't overclock, so I just assumed this was normal.

[L'enFer]

you did what many people were waiting for. thanks a lot.
could you tell me whether you test Scythe Infinity with 2 fans in push-push configuration?

[Spawne32]

please consider adding overclocked and overvolted results to your charts...

I don't really care about stock speed/volts... I want to see how the heatsinks handle under pressure... that TEC sink isn't gonna be the best performer when you add vCore

SMP folding (Linux SMP client) is designed for 4 cores... not sure about the Windows SMP client because it is unstable and inefficient compared to the Linux version so I've never bothered with it

The overclocked results will be added on the next set of tests, im gona take the top 5 heatsinks off of the charts, and pit them against each other with the maximum heat load during overclocking That should be somewhere above 3ghz with 1.5v

Spawne, isn't that Coolink heatsink exactly the same as in Noctua, the only difference being the fan? Oh, and less heatpipes too.

vapor thinks it is a similar design, only thing i can tell you is rascom does both noctua and coolink

you did what many people were waiting for. thanks a lot.
could you tell me whether you test Scythe Infinity with 2 fans in push-push configuration?

push/push as in both fans pushing into the heatsink in apposing fashions?

[L'enFer]

push/push as in both fans pushing into the heatsink in apposing fashions?

yes, 2 fans placed like this: fan-1>>heatsink<<fan-2. ">>"/"<<" is airflow direction.

[Spawne32]

yes, 2 fans placed like this: fan-1>>heatsink<<fan-2. ">>"/"<<" is airflow direction.

Thats an interesting suggestion, ill have to add that into my list.

[L'enFer]

Spawne32, yes, please do it. it's very interesting for me to know how good is Infinity in comparison with present "aircooling kings".