Noctua NH-D15S vs. Noctua NH-U14S

[Trinimini]

Greetings,

I've begun to start looking into overclocking my i7-4770k, and I think I've narrowed down my choices to Noctua's D15S and U14S. I would have liked the D15, but I don't think my RAM (Patriot Viper) would allow it. In a perfect world, I'd like to overclock to 4.5GHz (or as close as possible), but as long as I can reach 4.1GHz at least, I'll be happy. The D15S is currently $5 more than the U14S where I live (Canada). Taking into consideration temperature, sound level, price, and my overclock goal (4.1-4.5GHz), which cooler should I go for?

Thanks for the help.

 

[Darkbreeze]

I have the U14S on my Skylake i7-6700k with a 4.5Ghz OC and it does perfectly fine. Great in fact. It also has no RAM clearance issues on my Gigabyte Gaming 5 board and weighs a lot less than the D15 so that means less stress to the socket and motherboard. Unless you're planning a seriously over the top overclock, I'd just go with the U14S.

If you have tall RAM, I'd avoid the D15 as well, although there is a D15 model with better RAM clearance, but it's a single fan model and probably isn't terribly better than the U14S.

http://www.amazon.ca/dp/B00XUVGLEU/?tag=pcp0f-20

 

[Redneck5439]

Greetings,

I've begun to start looking into overclocking my i7-4770k, and I think I've narrowed down my choices to Noctua's D15S and U14S. I would have liked the D15, but I don't think my RAM (Patriot Viper) would allow it. In a perfect world, I'd like to overclock to 4.5GHz (or as close as possible), but as long as I can reach 4.1GHz at least, I'll be happy. The D15S is currently $5 more than the U14S where I live (Canada). Taking into consideration temperature, sound level, price, and my overclock goal (4.1-4.5GHz), which cooler should I go for?

Thanks for the help.

I just installed a NH-D15S and I love it. I made it into an even bigger beast by installing two additional fans (3 total) all in push/pull. I installed 2 3400rpm Delta fans and my CPU is cooled just as good as water cooling. The only way you would have any problems with the NH-D15S is if you installed additional fans to it. The heat sink is raised higher than the NH-D15 to allow for higher fined RAM to fit. If you really need to have additional fans you could always install 120mm fans and move them up slightly to clear components.

Really you can't go wrong with either the NH-D14S or the NH-D15S, but the NH-D15S does have better cooling capabilities and has been modified from the NH-D-15 to give high finned RAM clearance.

 

[Darkbreeze]

For a 4.5Ghz overclock, the U14S is plenty. Going over that probably warrants something bigger, but not necessarily. I don't exceed 65°C running Prime95 regardless of how long I run it. The same applied to my 8320 when using that cooler at a 4.5Ghz OC, and that was a 125w TDP chip. Your 4770k is an 84w TDP chip.

I could probably pull 4.7Ghz with this Skylake chip and still be comfortably within spec considering I have yet to exceed 63°C and Tjunction is 68°C with TJmax (Throttle) temps not being until 100°C. Of course, anything over 80°C is hotter than sustained usage should ever see since thermal fatigue is likely to begin to accrue when run at those temps consistently.

For your chip, Tjunction is 79°C with TJmax being 100°C. Any of the mentioned coolers will keep you well below those specs with a 4.5Ghz overclock unless you have been unlucky enough to get the worst binned chip possible.

Noteworthy is the fact that my results have been using a single fan on the U14S in a push configuration, and I've verified an additional 3°C drop in temps when using it with dual fans in a push pull configuration, by mounting a Noctua NF-A14 on the backside. The NF-A14 PWM has a higher overall CFM and static pressure ratio than the NF-A15 PWM that comes with the U14S which is recommended so that the rear "pull" fan isn't creating a resistance to the airflow of the front "push" fan, which is what would happen if an identical or lower CFM fan were mounted on the rear.

Static pressure doesn't matter much for the rear fan though since there is really no resistance to it's ability to move air since it's not forcing it through the heatsink like the front fan. The addition of an NF-A14pwm would be a great upgrade to the D15S if you decided to go that route and wanted to add an additional fan to it on the rear.

 

[Trinimini]

For a 4.5Ghz overclock, the U14S is plenty. Going over that probably warrants something bigger, but not necessarily. I don't exceed 65°C running Prime95 regardless of how long I run it. The same applied to my 8320 when using that cooler at a 4.5Ghz OC, and that was a 125w TDP chip. Your 4770k is an 84w TDP chip.

I could probably pull 4.7Ghz with this Skylake chip and still be comfortably within spec considering I have yet to exceed 63°C and Tjunction is 68°C with TJmax (Throttle) temps not being until 100°C. Of course, anything over 80°C is hotter than sustained usage should ever see since thermal fatigue is likely to begin to accrue when run at those temps consistently.

For your chip, Tjunction is 79°C with TJmax being 100°C. Any of the mentioned coolers will keep you well below those specs with a 4.5Ghz overclock unless you have been unlucky enough to get the worst binned chip possible.

Noteworthy is the fact that my results have been using a single fan on the U14S in a push configuration, and I've verified an additional 3°C drop in temps when using it with dual fans in a push pull configuration, by mounting a Noctua NF-A14 on the backside. The NF-A14 PWM has a higher overall CFM and static pressure ratio than the NF-A15 PWM that comes with the U14S which is recommended so that the rear "pull" fan isn't creating a resistance to the airflow of the front "push" fan, which is what would happen if an identical or lower CFM fan were mounted on the rear.

Static pressure doesn't matter much for the rear fan though since there is really no resistance to it's ability to move air since it's not forcing it through the heatsink like the front fan. The addition of an NF-A14pwm would be a great upgrade to the D15S if you decided to go that route and wanted to add an additional fan to it on the rear.

Even though the U14S would be more than enough for a 4.5GHz overclock, does the D15S (in single-fan mode) offer theoretically better temperatures, an even higher overclock, and comparable sound levels? As it stands currently, about $2 separates the two on Amazon, so if I can expect better performance with the D15S, I'm okay spending the extra $2.

Appreciate the info so far though.

 

[Redneck5439]

The NH-D15S outperforms the NH-D14S so it will allow for higher overclocks depending on how you did in the "chip lottery" with binning.

 

[Darkbreeze]

With a single fan, it will have "similar" sound qualities, since they both use the NF-A15 PWM fan. If you go with a dual fan model or add a second fan, of course Sound pressure levels will be higher.

Frostytech shows a 2°C difference in max temps between the regular D15 dual fan model and the NH-U14S, with a max 2.5db sound pressure difference.

http://www.frostytech.com/articleview.cfm?articleid=2766&page=5


Since there is almost a 11°C difference in max Delta temps between the NH-D15 and D15S, and the NH-U14S is within 2°C Delta temp as the regular D15, I find it hard to imagine that the D15S outperforms or offers a significant advantage over the U14S unless you add a second fan. Maybe not then. The D15S is actually one of the few coolers made by Noctua that I have no understanding for why it was even made.

http://www.eteknix.com/noctua-nh-d15s-cpu-cooler-review/5/

At max fans, the U14S has less than a 1°C difference in overclocked max delta as the Cooler master Seidon 240m liquid cooler, which is pretty amazing for a single fan, single width tower air cooler.

http://www.eteknix.com/noctua-nh-u14s-cpu-cooler-review/4/


Still, you probably can't go wrong with any of these coolers, but for myself, I prefer to limit how much weight stress and overall sound pressure level there is going to be, which is why I like the smaller Noctua cooler, and it still does a very admirable job. Your overclock is probably going to be more likely to be limited by it's inherent properties than it is by whether you choose one or the other of these coolers.

 

[Darkbreeze]

The NH-D15S outperforms the NH-D14S so it will allow for higher overclocks depending on how you did in the "chip lottery" with binning.

There is no Noctua NH-D14S, so that was either a typo or you are mistaken.

 

[Redneck5439]

The NH-D15S has an advantage over the NH-U14S. For one thing the NH-D15S, like the NH-D15 and NH-D14 is a dual tower cooler. The NH-U14S is a single tower cooler. The size of the heat sink allows for better cooling capabilities. A really good review of the NH-D15S worth reading:

http://www.tweaktown.com/reviews/7301/noctua-nh-d15s-style-cpu-cooler-review/index.html

Especially note where they test the thermals of the NH-D15S vs its competition.

 

[Redneck5439]

The NH-D15S outperforms the NH-D14S so it will allow for higher overclocks depending on how you did in the "chip lottery" with binning.

There is no Noctua NH-D14S, so that was either a typo or you are mistaken.

Typo 😉

Again though, you really can't go wrong with either one, they are both superior coolers giving you near water cooling (or equal depending on fan setup) without the worries of pump failures and coolant leaks.

 

[Darkbreeze]

Tweaktown also tested the NH-U14S, on their very same test platform, and the NH-U14S outperforms the NH-D15S by .25°C. That's why I would be hesitant to go with the cooler unless I was going to add a second fan. Clearly the size of the heatsink doesn't offer a great advantage over the single width U14S heatsink unless you also have an additional fan to help remove the heat that the larger heatsink is able to absorb. Absorption of heat isn't terribly helpful unless you can also dissipate or exchange it to the air and then out of the case.

Using two fans, the NH-U14S has a lower overclocked max temp (68°C) than the NH-D15 (69.3°C) that DOES have two fans. So given that, I'm not sure I'd ever want the heavier, more expensive cooler. And yes, those are both LGA 115x Intel socket test results. Even if there is some small amount of variance between tests, you're clearly in the same ball park with both coolers.

 

[pfilly]

Ran into this thread doing a google search because I was wondering the same thing about these two coolers and the best given answer is misleading; tests for the D15S are with a Haswell 4770K (known to run extremely hot and tests for the U14S are with the 2600K which is 32nm which is known to run much cooler. I almost cancelled my order for the D15S but had to dig a little to see why the results seemed too good to be true.

 

[damien1.english]

Ran into this thread doing a google search because I was wondering the same thing about these two coolers and the best given answer is misleading; tests for the D15S are with a Haswell 4770K (known to run extremely hot and tests for the U14S are with the 2600K which is 32nm which is known to run much cooler. I almost cancelled my order for the D15S but had to dig a little to see why the results seemed too good to be true.

holy necro but truth be told between the u14s and d15s the temps are maybe 2c +- on a 150 watt chip with just one fan on each but it wont really beat the d15s but the results of it being about the same are not that far off in dual fan config the u14s is quite the monster for its size http://www.relaxedtech.com/reviews/noctua/nh-d15/8 even with this being an older result. the differences between them wont really crop up until you hit much higher wattage chips then say an 8700k. I switched from phanteks tc14pe that I was running in single fan to a u14s and temps are practically the same on my 5ghz 8700k. but if you have the room go for it as the price diff is very minimal

 

[Saga Lout]

You're right about the necro, damien1.english, and that's why I'm going to close the thread now. I'll leave your link in but it's even older than this thread so there is little to be drawn from either.

Your post will still be seen and read but there can be no more relevant replies this far down the line.