be quiet! Dark Rock Pro 2

[crazyeyesreaper]

be quiet! looks to dominate the CPU-cooling market with their Dark Rock Pro 2, having armed this behemoth with two SIlentwing fans, dual cooling towers, and seven 6 mm heatpipes. But whether it will live up to the be quiet! name or falters in the face of stiff competition remains to be seen.

Show full review

 

[Lauen]

The second language is german, not dutch

 

[Farmer Boe]

Great review! Nice to see how this cooler stacks up since they are becoming more widely available in North America.

 

[crazyeyesreaper]

Yea The Install hardware is a bit fiddly but the Dark Rock Pro 2 is definitely a top notch air cooler.

 

[AsRock]

Yea The Install hardware is a bit fiddly but the Dark Rock Pro 2 is definitely a top notch air cooler.

Sadly with a top notch price tag.. Nice review btw.

 

[Sasqui]

Sadly with a top notch price tag.. Nice review btw.

I'm pretty stunned, anyone on the fence about turn-key water cooler (hint: Corsair H110) vs air should definitely consider this. I say it's worth the premium considering reliability, and less install hassle than water.

 

[buildzoid]

what mounting holes does the 135mm fan. Are they 120mm?

 

[Johan]

Given that the Noctua U12S is already very competitive with this thing with higher temps but lower noise in general, I was wondering how the U14S, which has been reviewed recently, would do.

But it's gone. Will it get tested on the new testbed any time soon? And maybe lower the testbed change frequency to something like once every 1.5~2 years or so. Using a different heat source shape shouldn't change the relevance that much...

 

[Ed_1]

I second that, Noctua U14S looks to be one of best for single tower HS and still quiet .

Question, why was stock clock voltage static 1.15 ?

Those max AIDA max temps are high , do AIDA FPU run hotter than linX ,Intel burn ?

 

[crazyeyesreaper]

The Noctua NH-U14S also has the issue of where it potentially can block the Top Expansion slot however so keep that in mind. On certain high end boards that are mostly PCIe focused this can become a problem for Multi-GPU systems as well as entry lvl systems that may only feature 1 PCIe 16 and 1 PCIe x4

And the fact of the matter is that Haswell doesnt really produce that much heat.

due to CPU DIE > TIM > IHS > TIM > Heatsink if the CPU is delidded temps would drop by a large amount and temp difference would be greater between coolers however the average user wont be delidding a chip.

AIDA FPU is roughly the same as Prime95 etc They all vary but AIDA has proved to be reliable

Stock voltage of 1.15 volts is due to how Haswell was designed as Stock voltage under AVX loads the CPU receives a voltage boost of 0.10

so in AIDA64 / Prime95 / IBT etc the cpu at stock would get 1.23-1.26 volts depending on the CPU where as all other loads are around 1.06-1.15v

For my CPU to be stable at 4.7 GHz it only needs 1.25 volts for those tests aka this means stock the system applies extremely high voltage and skews the results as the voltage increase was no necessary. As such with the overclocked voltage being lower this means stock temps were getting higher than overclocked temps due to this issue which only happens in AVX stress tests. In every other test including day to day operation the voltage is lower thus Temperature testing makes no sense. I honestly hate haswell as a platform for testing but it is the mainstream platform going forward. To put it bluntly for testing a cooler and its ability to handle heat the SB-E platform was superior.

That and using adaptive or offsets to can cause voltages in these AVX tests to get out of hand when they are perfect for regular tests. With offsets I have seen volts jump as high 1.32 volts. Thus static voltages were used to give even stable results that can be easily compared. As this removes a potential variable that could influence tests and outcomes.

 

[AsRock]

I'm pretty stunned, anyone on the fence about turn-key water cooler (hint: Corsair H110) vs air should definitely consider this. I say it's worth the premium considering reliability, and less install hassle than water.

You have a point although i think water coolers for what they are are expensive. I must admit that this cooler would look dam sexy with some Cougar fans on it.

 

[crazyeyesreaper]

I think I will use this cooler on my SB-E rebuild or my gaming system

Performance / Silence / Looks

it has everything I want lol.

 

[Ed_1]

The Noctua NH-U14S also has the issue of where it potentially can block the Top Expansion slot however so keep that in mind. On certain high end boards that are mostly PCIe focused this can become a problem for Multi-GPU systems as well as entry lvl systems that may only feature 1 PCIe 16 and 1 PCIe x4

And the fact of the matter is that Haswell doesnt really produce that much heat.

due to CPU DIE > TIM > IHS > TIM > Heatsink if the CPU is delidded temps would drop by a large amount and temp difference would be greater between coolers however the average user wont be delidding a chip.

AIDA FPU is roughly the same as Prime95 etc They all vary but AIDA has proved to be reliable

Stock voltage of 1.15 volts is due to how Haswell was designed as Stock voltage under AVX loads the CPU receives a voltage boost of 0.10

so in AIDA64 / Prime95 / IBT etc the cpu at stock would get 1.23-1.26 volts depending on the CPU where as all other loads are around 1.06-1.15v

For my CPU to be stable at 4.7 GHz it only needs 1.25 volts for those tests aka this means stock the system applies extremely high voltage and skews the results as the voltage increase was no necessary. As such with the overclocked voltage being lower this means stock temps were getting higher than overclocked temps due to this issue which only happens in AVX stress tests. In every other test including day to day operation the voltage is lower thus Temperature testing makes no sense. I honestly hate haswell as a platform for testing but it is the mainstream platform going forward. To put it bluntly for testing a cooler and its ability to handle heat the SB-E platform was superior.

That and using adaptive or offsets to can cause voltages in these AVX tests to get out of hand when they are perfect for regular tests. With offsets I have seen volts jump as high 1.32 volts. Thus static voltages were used to give even stable results that can be easily compared. As this removes a potential variable that could influence tests and outcomes.

Ok, but isn't that the case with all HS with 140mm fans compared to 120 . NH-u14S doesn't look much wider than fan (150mm ? ).
my P8Z77 has PCI-e 1x ? , small slot at top so guess that would be useless .

Ok, I forgot about certain instructions Haswell boost the voltages , I thought they run slightly less than ivy bridge (1.00-1.05 or so ) .
Thanks

 

[crazyeyesreaper]

Depends on the cooler

Example the Phanteks PH-TC14PE or w.e is huge uses 140mm fans no issues blocking top expansion slot. So it depends on the Cooler While I like the Noctua NH-u14S and U12S the U12S is the better overall product in my eyes thus why its in the charts. however both coolers with dual fans cost as much or more than the Phanteks PH-TC14PE / be quiet Dark Rock Pro 2 which both go on sale rather frequently into the high $80 low $90 price range while the U14S can cost more than $100 with a second fan. So it really does depend on the end user.

Again depends on the CPU crappy chips tend to have higher voltages in the 1.10+ range while good chips are 1.0 or lower. My CPU has lower voltage so it clocks rather well but for clarification there are many users on these forums that need 1.25-1.3 volts just to get to 4.3-4.4 GHz which drastically alters heatload and cooling efficiency.

BUt as always there is the fact of diminishing returns to remember. It applies to CPUs / GPUs / Coolers etc the higher the price you can usually expect better performance but to gain a few % points you might have to spend quite a bit. Its the sad reality. But all things considered the quality of be quiet!'s Dark Rock Pro 2 on top of its low noise levels and performance I would take it over the NH-U12S or U14S especially when you consider the fact the be quiet! is the better performer while being quieter for the same price.

100% fan speed tests can be compared between SB-E and Haswell cooler reviews and when paired with 2 fans the U14S is louder but this is subject to user interpretation as well because some are happy with loud fans others are not it all a personal preference thing so i just try to give repeatable results and my experiences with the product.

 

[Fourstaff]

This little monster at £56 is cheaper than the Phanteks on this side of the pond, and cheaper than the U14S too. Worth considering if I haven't already have the H100i

 

[Ed_1]

Depends on the cooler

Example the Phanteks PH-TC14PE or w.e is huge uses 140mm fans no issues blocking top expansion slot. So it depends on the Cooler While I like the Noctua NH-u14S and U12S the U12S is the better overall product in my eyes thus why its in the charts. however both coolers with dual fans cost as much or more than the Phanteks PH-TC14PE / be quiet Dark Rock Pro 2 which both go on sale rather frequently into the high $80 low $90 price range while the U14S can cost more than $100 with a second fan. So it really does depend on the end user.

Again depends on the CPU crappy chips tend to have higher voltages in the 1.10+ range while good chips are 1.0 or lower. My CPU has lower voltage so it clocks rather well but for clarification there are many users on these forums that need 1.25-1.3 volts just to get to 4.3-4.4 GHz which drastically alters heatload and cooling efficiency.

BUt as always there is the fact of diminishing returns to remember. It applies to CPUs / GPUs / Coolers etc the higher the price you can usually expect better performance but to gain a few % points you might have to spend quite a bit. Its the sad reality. But all things considered the quality of be quiet!'s Dark Rock Pro 2 on top of its low noise levels and performance I would take it over the NH-U12S or U14S especially when you consider the fact the be quiet! is the better performer while being quieter for the same price.

100% fan speed tests can be compared between SB-E and Haswell cooler reviews and when paired with 2 fans the U14S is louder but this is subject to user interpretation as well because some are happy with loud fans others are not it all a personal preference thing so i just try to give repeatable results and my experiences with the product.

Ok, thanks for clearing that up . I take it from pictures of room clearance on the DR2 and PH-TC14PE . that memory with small-mid height would work , like mine Crucial Ballistix Tactical Tracer which heatsink are only 1/4"/6mm above PCB ?
Crucial Ballistix Tactical Tracer 8GB (2 x 4GB) 24...

 

[crazyeyesreaper]

Yes those should fit or at least you should be able to position the fan a bit higher to make room. Anything taller than that would be problematic so with these coolers low profile memory is definitely a necessity. Sticks like Gskills Sniper series may fit but I do not have a set on hand.

But yes as far as your Ballistix Tactical Tracers are concerned you should experience no issues.

 

[Ed_1]

thanks again .
This HS seem pretty high cost , I only see it at amazon for 108$

 

[crazyeyesreaper]

NCIX.com has it for both canada and usa last i checked it was on sale for $89 a few days ago

 

[Ed_1]

Thanks, I am bookmarking that site, some good prices .
They have it listed for 69.99$ now .

 

[Pumper]

Bought this cooler and am choosing a mobo for 4770K (upgrading from X6). I am looking at Asus Gryphon, but some people say that big heatsinks on it block the top PCI-E slot. Will I have problems with Dark Rock Pro 2 + 4770K + 660Ti DC2? Thanks.

 

[crazyeyesreaper]

what motherboard exactly in terms of model number

 

[Pumper]

GRYPHON Z87 (or 90MB0EH0-M0EAY0)
ASUS GRYPHON Z87 LGA 1150 Intel Z87

 

[crazyeyesreaper]

It might have clearance issues but what i can tell you is its not the cooler but the fan clips with a bit of ingenuity you can bend them slightly. I dont think it will be neccessary however but regardless it will be a tight fit. But if there is an issue it can be easily overcome.

 

[Pumper]

Installed the new components yesterday. Did not have any issues with clearance - still get 3-4mm of space between the fan clips and GPU - over 1cm between heatsink and GPU.

Idling at ~25C with 20C room temperature. Will be overclocking to see that my 4770K can manage (default voltage is at 1.04v in BIOS and is showing 1.109v when running Cinebench).