Noctua NH-D14 vs Corsair H100 for 2600K OC'ing? Update: Went with both!

[stahlhart]

The review you linked has been questioned all over the Internet. It's one of the few reviews out of 15 or so where the SA lost. Otherwise, the SA beats the NH-D14 by 1-2*C, while also being significantly quieter.

It's hard to believe he had problems with the 950 i7 (4C) @ 4.4ghz @ 1.45V when the SA holds the world record for air cooling by getting a Core i7 970 (6C) to 5.0ghz at 1.52V.

Tom is highly biased in general in favor of the Noctua and against the Thermalright, for reasons that go above and beyond the performance tests. It really shows at the end here:

http://www.youtube.com/watch?v=07XTN0Qll2o

 

[Idontcare]

I decided to just screw it, bought a tube of NT-H1 and a NH-D14 for good measure. I try all combos. H100 and NH-D14. Stock TIM, NT-H1, AS5, TX-2, Ceramique, and I've got one other one I can't recall the name of at the moment (might be MX-2).

We'll see how many combos I have the patience to test. I'm also a lapper too, I'll do stock unpolished surfaces for the first rounds but eventually I will be lapping these things too.

(dang, I can already feel the itch to drag out my trusty vapoLS too...not sure how I would get it to mate with a LGA1155 socket though)

 

[RussianSensation]

I decided to just screw it, bought a tube of NT-H1 and a NH-D14 for good measure. I try all combos. H100 and NH-D14. Stock TIM, NT-H1, AS5, TX-2, Ceramique, and I've got one other one I can't recall the name of at the moment (might be MX-2).

Oh snap! This calls for a Highest End CPU cooler vs. Highest End Corsair Watercooling system Roundup thread + TIM roundup.

 

[Spikesoldier]

I decided to just screw it, bought a tube of NT-H1 and a NH-D14 for good measure. I try all combos. H100 and NH-D14. Stock TIM, NT-H1, AS5, TX-2, Ceramique, and I've got one other one I can't recall the name of at the moment (might be MX-2).

We'll see how many combos I have the patience to test. I'm also a lapper too, I'll do stock unpolished surfaces for the first rounds but eventually I will be lapping these things too.

(dang, I can already feel the itch to drag out my trusty vapoLS too...not sure how I would get it to mate with a LGA1155 socket though)

yeah, now youre obligated to do a ATF review comparison.

expecting your thread

 

[ViviTheMage]

So which was better?

 

[Idontcare]

So which was better?

Still setting up.

Waiting on my IR gun and my sound meter.

I'm off to Ocean City MD this Saturday and won't be back for a week, so realistically it'll be another 2 weeks before I'll be updating the thread with data.

 

[Helios1234]

I spent some time reading reviews and realized that most reviews tend to compare the coolers at Low, Medium or High fan speeds. The reviewers then proceed to simply list out the dB and temperatures in a chart. However, that tells us little about how comfortable those fan settings are (unless you can easily tell the difference between 40 vs. 42 vs. 48 dB, etc.).

Finally, I found this video review which limited the top air coolers to only 40 dB using a fan controller: http://www.youtube.com/watch?v=YLJxEWPrEDs

Fast forward to 9min:30sec to see that 40 dB at 15 cms is pretty quiet. This drops to only 31.5 dB 1 meter away.

The testing involved a Core i7 920 @ 4.3ghz with 1.42V:

All of the coolers were equipped with 2x140mm fans, limited to 40 dB using a fan controller, MX-3 thermal paste was used for consistency:

Thermalright Silver Arrow = 84*C
Noctua NH-D14 = 92*C
Prolimatech Megahalems Super Mega = 95*C
Thermolab BARAM 2010 = 97*C

Another great cooler they didn't test though was the Thermalright Archon with 2 x TY-140mm fans.

As you can see, the Noctua NH-D14 which performed extremely well in most reviews such as Vortez is actually louder than the Silver Arrow. So once its fan speeds are dropped to about 950-1000 rpms (40 dB), it is no longer able to maintain its leadership positioning.

I myself just grabbed the Silver Arrow at Mwave.com for $70. Just call them and ask for free shipping (since they were offering it last week and say that you missed it by a couple of days). I don't doubt that the H100 is the more powerful cooler (esp. in a push/pull config). However, at $120, it's $50 more expensive, and it appears that it achieves the best performance at much higher noise levels based on that HardOCP review.

As an added bonus, SA ships with Thermalright Chill Factor III, which is one of the better thermal pastes.



If you are ok with the added noise levels, a 3x120mm radiator water cooling system will be about 10*C better than the top air coolers. However, things change completely once you start comparing them at similar noise levels:

"If we compare our today’s hero, Swiftech H2O-X20 Edge, against the today’s best air cooler [Thermalright Archon], we have to keep in mind the acoustic performance of both these systems and we are going to get to it in the next chapter of our article. However, even now it already becomes obvious that in its quietest mode with all fans at 950 RPM and the pump at 2150 RPM Swiftech H2O-X20 Edge is only 2°C better than Thermalright Archon with its default fan TY-140 at 1260 RPM." - Xbitlabs

The H2O-X20 costs $320! You can add another fan to the Archon and you'll have a very quiet air cooler within a hair of that water cooling system for 3x less cost.

In my review (Vortez one), I did highlight the fact that the TR-140 fans are quieter than the Noctua ones. In fact, I always recommend the Silver Arrow over the Noctua for the reasons you gave. It's quieter, cheaper, and comes with the TF3 thermal paste which is amongst the best.

There was actually another review from a big review site that showed the SA greatly lagging behind the D14 when both were pushed to their limits...to the point were the SA was no longer safe for the CPU. The author stated that he repeated the tests several times and got the same results each time.

Edit: here you go:

http://www.overclock3d.net/reviews/cases_cooling/thermalright_silver_arrow_vs_noctua_nh-d14/1

At 4.2 GHz, the Silver Arrow is slightly edging out the NH-D14. However, at 4.4 GHz, the Silver Arrow seems to hit a performance wall where the NH-D14 does not. The test was repeated on an open bench and inside a case, with the results being the same.

Tom's review is flawed and you can see how biased he is. I remember him asking me for help because he couldn't install it properly and simply made the video to get hits and more attention because at the time, the Noctua VS Silver Arrow competition was hotting up. What is suspicious is the way he awards the Silver Arrow a lower award despite the fact that it actually beats the Noctua in two of his OWN tests. This is the reason why most don't see him as a credible source on the matter.

Xbitlabs on the other hand provide a much better review and one that compliments what others have found very well.

 

[darckhart]

I have the h100 on a 2600k, and I would say fan setting 2 and higher is already considered too loud for "normal" people (e.g. roommates). However, even at setting 1 and oc to 4.5 GHz, the temps are good. (using IC7 instead of their default tim.) I went with h100 simply because I don't want 900 g (1200 w/ fans) of heatsink on my cpu and mobo.

 

[RussianSensation]

In my review (Vortez one), I did highlight the fact that the TR-140 fans are quieter than the Noctua ones. In fact, I always recommend the Silver Arrow over the Noctua for the reasons you gave. It's quieter, cheaper, and comes with the TF3 thermal paste which is amongst the best.

I am actually surprised that so few hardware enthusiasts choose the Silver Arrow over the NH-D14. Perhaps a part of it has to do with Noctua's superior Ram clearance with taller heatsinks:

1.

2.

I have the h100 on a 2600k, and I would say fan setting 2 and higher is already considered too loud for "normal" people (e.g. roommates). However, even at setting 1 and oc to 4.5 GHz, the temps are good. (using IC7 instead of their default tim.) I went with h100 simply because I don't want 900 g (1200 w/ fans) of heatsink on my cpu and mobo.

Even the original Tuniq Tower 120 was already pushing 800 g without a fan. Modern CPU heatsinks such as the Tuniq Tower Extreme, Archon, NH-D14, Megahalems, Ultra 120 Extreme / Venomous X, Havix 140 all push nearly 1-1.3 kgs with fans. I don't recall ever hearing a story of anyone damaging their motherboard with such heatsinks if it's secured with a metal backplate. In other words, your fears are largely unfounded. It wouldn't be recommended to ship a system with UPS with such a heatsink mounted in transit however.

Think about this, a reference HD6970 card weighs about 1.1 lbs (500 grams). And people run 3 of those in Tri-Fire on a motherboard without problems. So if 1.5 kgs of weight doesn't damage the motherboard (and that's without a backplate), then surely 1.2-1.3 kgs with a backplate is not a problem.

 

[stahlhart]

I am actually surprised that so few hardware enthusiasts choose the Silver Arrow over the NH-D14. Perhaps a part of it has to do with Noctua's superior Ram clearance with taller heatsinks:

Corsair has apparently addressed this with low-profile versions of the Vengeance DIMMs.

I've also been toying with the idea of going with one (or two) Gentle Typhoons in place of the stock fans, to add memory options for the build... but the TY-140s are just working so darn well that I can't bear the thought of replacing them.

 

[darckhart]

I don't recall ever hearing a story of anyone damaging their motherboard with such heatsinks if it's secured with a metal backplate. In other words, your fears are largely unfounded. It wouldn't be recommended to ship a system with UPS with such a heatsink mounted in transit however.

Think about this, a reference HD6970 card weighs about 1.1 lbs (500 grams). And people run 3 of those in Tri-Fire on a motherboard without problems. So if 1.5 kgs of weight doesn't damage the motherboard (and that's without a backplate), then surely 1.2-1.3 kgs with a backplate is not a problem.

To each his own, of course. I'm simply throwing out 1 of the reasons I went with h100. (and having used enough air coolers and custom wc over the years, i find the h100 to be a decent value.) (Also, video cards are mounted in the middle of the board, not subject to causing the same cantilever action.) In any case, the NH-U12P (about 800 g) on my LGA775 system has definitely bowed, drooping down 0.5 cm over the years. Whether or not that has "damaged" anything was not my concern; simply having caused the drooping was enough for me.

 

[gramboh]

I've been thinking about H100 vs. Silver Arrow vs. D14 for my next build, very happy with the U120 Extreme on this old system, so I'd lean toward the SA, but the weight and size of these buffers is a bit insane. I'd consider spending the extra $$ on the H100 but I don't like the talk of loud fans. Do the rads really need that much airflow?

 

[Axon]

I find the h100 fans to be tolerable on the balanced setting, and the performance on that level is more than sufficient. I previously had a Megahalems in push/pull, and my load temps were low 80s. h100 load is mid to high 60s. If you're already enjoying good temps, the h100 will essentially be a side grade. But if you feel your temps are high, it should be an upgrade.

Additionally, you can always throw custom fans on there. I have a pair of AP-15s on my h100 and I don't hear much except on high; even then, not so bad. No real difference in performance.

Finally, you can attach the fans to an external fan controller for much easier adjustment.

 

[RussianSensation]

I previously had a Megahalems in push/pull, and my load temps were low 80s. h100 load is mid to high 60s. If you're already enjoying good temps, the h100 will essentially be a side grade. But if you feel your temps are high, it should be an upgrade.

Wow, that's a serious performance upgrade considering the Megahalems with push/pull is easily in the top 5 best air coolers!

It's surprising your temps with the 2600k @ 4.6ghz were in the low 80s with 2 fans on the Mega. My i7 860 @ 3.9ghz (which consumes way power than the 2600k @ 4.6ghz), ran at 62-65*C with just 1 x 1600 rpm fan with all 8 threads loaded.

 

[dakata24]

Curious what the verdict was between the H100 vs Noctua?

The cpu cooler is the last piece to my latest i7 build. Torn between either the H100 or go with air cooling with either Thermalright Silver Arrow, Archon, or Noctua.

 

[RussianSensation]

Curious what the verdict was between the H100 vs Noctua?

The cpu cooler is the last piece to my latest i7 build. Torn between either the H100 or go with air cooling with either Thermalright Silver Arrow, Archon, or Noctua.

The Noctua is $69.99
The Havik 140 is $59.99
The Silver Arrow is $69.99
The H100 is $100

Here is a review:
http://www.overclock3d.net/reviews/cases_cooling/corsair_h100_review/1

Core i7 950 @ 4.2ghz

Delta between Ambient and Load temp:

Noctua NH-D14: 50.45*C
Silver Arrow: 49.6*C
H100 (low fan speed): 55.9*C
H100 (high fan speed): 49.5*C

Core i7 950 @ 4.4ghz

Noctua NH-D14: 67.7*C
H100 (high fan speed): 60.1*C

"So when testing at 4GHz the H100 was just over 2c warmer than the NH-D14, that was the only test it lost though. I say that in honest shock and I did retest each one several times to make 100% certain. So yes as the graphs suggest when running the H100 at maximum in both the 4.2GHz and 4.4GHz the H100 was cooler than the NH-D14 and only beaten by full on watercooling kits. The trade off of these temperatures is some quite high noise levels, indeed it's far from silent."

So if you want the ultimate performance regardless of noise levels (or cost: i.e., H100 with 4 fans), then H100 is the way to go. Otherwise, at similar noise levels, its performance isn't that different for $30-40 more.

In other words, any of these coolers will let you take your 2500k/2600k to 4.8ghz, and anything beyond that is probably going to require unsafe voltage levels anyway.

The Havik 140 for $59.99 is an incredible value for a high-end air cooler. It was able to keep the i7-990X (6C/12T) @ 4.4ghz stable at 2x1200 rpms!

If you really want to go balls out, the RS360 kit is $147. For $50 more over the H100, you get 12-15*C lower temperatures.

But I mean considering the Havik was able to cool a 6c/12t 32nm chip, it should have no problem with your 2600k 4C/6T for $60.

 

[Yellowbeard]

So.....what was the conclusion by the OP?

 

[mv2devnull]

@Yellowbeard:
IDC mentioned two weeks delay about two weeks ago, so he hasn't really missed the schedule yet.

I am actually surprised that so few hardware enthusiasts choose the Silver Arrow over the NH-D14. Perhaps a part of it has to do with Noctua's superior Ram clearance with taller heatsinks:

By specs the NH-D14's fins are only about 3 mm higher than those of SA. What is does more is the size difference of the fan: 120 mm vs 140 mm.

In fact, if you take just a plain 30 mm high RAM and put the 140 mm TY-140 on top of it, where does the "can take a 165 mm cooler" case go? No such issue with stock fan NH-D14.

 

[Idontcare]

So.....what was the conclusion by the OP?

I got all my parts, now just need some quality time to spend doing the tests.

My 7yr old and 4yr old don't start school until after Labor Day, acoustic measurements will be dodgy at best until then.

 

[RussianSensation]

The cpu cooler is the last piece to my latest i7 build. Torn between either the H100 or go with air cooling with either Thermalright Silver Arrow, Archon, or Noctua.

Xbitlabs Review

"For example, the lowest noise from Corsair H100 with two fans rotating at their minimal speed is equal to the highest noise from Thermalright Archon. And the next two comparison points even farther apart. Therefore, going back to the results chart we can conclude that at about the same level of noise the best air cooler is 14°C more efficient than the new Corsair H100."

Looks like if you want the highest performance, H100 > top end air coolers, esp. with 4 fans. If you want a compromise between performance and lower noise levels, the air coolers are still the better option, and at a cheaper price of entry.

 

[blackened23]

I use the corsair H50 and it doesn't make any noise at all really - I love it. Now the H100 is louder than the NH-14 but only if you use the corsair (junk) fans. I haven't used the H100 personally but if the included corsair fans are like the ones that come with the H50....there's no way in hell i'd use the stock ones.

Nothing against the NH-14 of course, its definitely a great product if you dont mind the large case footprint.

 

[Helios1234]

@Yellowbeard:
By specs the NH-D14's fins are only about 3 mm higher than those of SA. What is does more is the size difference of the fan: 120 mm vs 140 mm.

In fact, if you take just a plain 30 mm high RAM and put the 140 mm TY-140 on top of it, where does the "can take a 165 mm cooler" case go? No such issue with stock fan NH-D14.

I'm not quite sure what you mean there. Even with the fan installed on top of the RAM, the overall height won't go above the heat pipes i.e. 160mm so its case compatibility is exactly the same as the Noctua's. Plus there is still the alternative of placing the front fan to the rear.

 

[RussianSensation]

I use the corsair H50 and it doesn't make any noise at all really - I love it. Now the H100 is louder than the NH-14 but only if you use the corsair (junk) fans.

But if you buy the H50/60/70/80/100 and proceed to replace its stock fans, then aren't you quickly eroding the value proposition? But then if you aren't aiming for top performance at low noise levels, what is the point of watercooling exactly? If you are only doing mild overclocks and want to spend $70-80, you can have a passively cooled HR-02 setup. I suppose if you absolutely dislike metal heatsinks and are aiming for aesthetics and the cool factor of watercooling, it's worth it. To me personally, the whole idea behind watercooling is better performance at the same noise level OR similar performance at a lower noise level that makes it superior to air cooling. Maybe I am underestimating how many people want the cleaner layout that Corsair's units offer.

 

[infoiltrator]

The problem is there are differences between open air and in case testing, and internal case arrangement.
Also a heat exchanger is a heat exchanger.
Fans, and fan noise, especially perceived fan noise, are very various.
Given size limitations all the"best" performers all things being equal will do about the same.
Dramatic changes in temp mean other issues are being addressed.
Thats the way I see it.

 

[blackened23]

But if you buy the H50/60/70/80/100 and proceed to replace its stock fans, then aren't you quickly eroding the value proposition? But then if you aren't aiming for top performance at low noise levels, what is the point of watercooling exactly? If you are only doing mild overclocks and want to spend $70-80, you can have a passively cooled HR-02 setup. I suppose if you absolutely dislike metal heatsinks and are aiming for aesthetics and the cool factor of watercooling, it's worth it. To me personally, the whole idea behind watercooling is better performance at the same noise level OR similar performance at a lower noise level that makes it superior to air cooling. Maybe I am underestimating how many people want the cleaner layout that Corsair's units offer.

Yeah, it does change the value proposition, no question. I think the noise problem is unique to the H100 though -- I haven't heard any similar such problems with the H50/H80. The H50 for me is dead quiet. As far as the price between the 2 units, they seem pretty close on most websites. Newegg and amazon usually have the NH-14 around 85-90 bucks -- and i'm hesitant to shop anywhere else for pc parts -- and i'm sure the H100 will come down in price since its a fairly new product.

Anyway, I guess the smaller footprint is a plus for me. My old cooler was so large that it was a pain to even change a video card in pci-e slot 1 (on a P8P67 board) and the NH-14 would protrude over the RAM slots. Kind of annoying. The NH-14 is an excellent product but damn...things huge.