Corsair vs Noctua

[pandasaurus]

Presently running Corsair ml140/120 pro fans inside an nzxt s340 elite tower, wondering if I’d see any change in temps/dB fan noise etc switching to NF-A14/12 3k rpm fans. My corsairs are ~2k rpms give or take.
Systems;
Tower: nzxt s340 elite
Motherboard: asus Maximus hero viii
CPU: Intel i7-6700k (OC'd to 4.8@1.295v)
Cooler: nzxt kraken x62 with ml140pro fans push/pull
GPU: asus nvidia strix RoG 1080Ti 11gb OC edition
Sound card: Strix asus soar
PSU: evga superNOVA g2 750w gold
RAM: 2x8gb patriot viper 4 2800mhz ram
Wifi card: asus pce-ac56
SSD: Samsung evo 850 as primary 840 as secondary
Lighting: NZXT hue+ with extension kit
Fan: ml140pro x1 top exhaust ml120pro X1 rear exhaust

 

[Darkbreeze]

I'm assuming you are referring to the Noctua Industrial series fans, since there are NO 3000rpm standard, chromax or redux series fans.

You will definitely NOT cut down on the noise level using those, but you will absolutely increase the cooling capability since those fans push about 158.5 CFM and have a whopping 10.5 mm H₂O static pressure. That's about double the CFM of the ML120 and about 2.5x the static pressure.

http://noctua.at/en/products/product-line-industrial/nf-a14-industrialppc-3000-pwm/specification


However, as a caveat, at identical RPMs the Noctuas will undoubtedly be quieter, it's just that they are unlikely to run at identical RPMs at the same thermal levels, unless you hamstring the fan using a voltage reduction adapter and that sort of defeats the purpose of using those fans.

Max decibel level though for the Corsair ML120's is about 35cfm, probably sounds more like 38cfm, and the Noctua PPC industrial 3000rpm fans have a max decibel level of 41.3db, so there's not a staggering noise difference between the two considering nearly double the cooling capability if not more considering the massive static pressure those fans have when used as intakes, on heatsinks or on radiators.

 

[pandasaurus]

Yeah I guess my big question is, if I swap out my ml140 pros in push pull on my kraken x62 as well as the top and rear exhaust case fans, will I see a noticeable change it temps?

 

[Darkbreeze]

I would think so, but it depends. If your current setup is already keeping your coolant temps at or near ambient, then you won't. Nothing short of active refrigeration of some kind can reduce temps below ambient, but I seriously doubt your current configuration comes anywhere close to that and clearly with double the airflow and 2.5x the static pressure, the Noctua Industrial fans will likely come at least closer to that goal.

How many case fans do you currently have and what models are they. Also, where are they installed and in what orientation (Intake, exhaust)?

 

[pandasaurus]

Presently I’m running a nzxt s340 elite tower fans in every place.. on the front for intake I have my x62 with 4 Corsair ml140 pros on it in push pull, a ml140 pro on the top rear for exhaust and a ml120 pro on the rear for exhaust, i idle at around 22c/73f, at the moment I’m getting some temp spikes but I think that’s cause I put a touch too much thermal paste when placing the x62 on. I just ordered 4x noctuas for push pull on my x62 for now, probably replace the others shortly down the road.

 

[pandasaurus]

Thank you for the help by the way, very much appreciated.

 

[Darkbreeze]

Idle temps don't make the least bit of difference, at all, to anything, unless the idle temps are like in the upper 40's or in the 50's. And if they are, then there is a problem.

Anything below 40°C at idle is fine. Of course, lower is better, but it really has no effect on what the load temps will be. Idle temps are largely related to what the ambient temperature is in the room, while load temps are strictly related to the cooling system, clock speed and voltage applied.

What are your load temps while running Prime95 version 26.6 on Small FFT option for 15 minutes?


I'm going out on a limb and say you will error out because I think your overclock is unrealistically high and your voltage is too low to be stable. I have one of the best 6700k's I've come across as judged by what I've seen 100's of others be able to achieve with theirs, and at 4.8Ghz I have to run like 1.4v to be stable. At 4.5Ghz I have to run 1.325v for stability. I'd bet you will find your OC is not stable running at that voltage. Run Prime95 and find out. It will also tell you if you are staying within the correct thermal envelope. If your temps go past 80°C while running Prime95, then you need to reduce your overcock because you sure as hell can't reduce the voltage any more.

Use HWinfo (NOT HWmonitor or Openhardware monitor, or any of the utilities that come with any of your hardware. They are pretty much all crap and are usually not accurate at all) to keep an eye on thermals and take screenshots of the CPU thermal sensors after about ten minutes, if it makes it that long. Post the results here. HWinfo or CoreTemp are the most accurate of all the utilities for monitoring that are out there. I actually like CoreTemp better, so it's up to you which one you use.

HWinfo: https://www.hwinfo.com/download.php


Core Temp: http://www.alcpu.com/CoreTemp/


Prime95 v26.6 is THE primarily accepted way to do the majority of baseline stability and thermal limit testing running the Small FFT option.

Prime95 version 26.6: http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html


Further, as explained to me and many others on this forum by Computronix, who has far more experience with CPU architectures and testing procedures than 95% of the people you will ever meet, speak to or read about. He is also the author of the Intel temperature guide, found here:

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html



This pretty well sums things up and is equally relevant whether working with an Intel or an AMD system.

I can think of several reasons why x264 encoding or AVX / AVX2 / FMA3 apps won't work as a unilateral metric for thermal testing.

(1) A steady-state workload gives steady-state temperatures; encoding does not.

(2) Simplicity in methodology; most users would find encoding apps unfamiliar and cumbersome to accomplish a simple task.

(3) Most users such as gamers never run any apps which use AVX / FMA, so adaptive or manual voltage aside, it makes no sense to downgrade your overclock to accommodate those loads and temps.

(4) Standardization; Prime95 has been around since 1996; many users are familiar with it.

For the minority of users who routinely run AVX / FMA apps, then P95 v28.5 can be useful tweaking BIOS for thermal and stability testing.


regardless of architecture. P95 v26.6 works equally well across all platforms. Steady-state is the key. How can anyone extrapolate accurate Core temperatures from workloads that fluctuate like a bad day on the Stock Market?

I'm aware of 5 utilities with steady-state workloads. In order of load level they are:

(1) P95 v26.6 - Small FFT's
(2) HeavyLoad - Stress CPU
(3) FurMark - CPU Burner
(4) Intel Processor Diagnostic Tool - CPU Load
(5) AIDA64 - Tools - System Stability Test - Stress CPU

AIDA64's Stress CPU fails to load any overclocked / overvolted CPU to get anywhere TDP, and is therefore useless, except for giving naive users a sense of false security because their temps are so low.

HeavyLoad is the closest alternative. Temps and watts are within 3% of Small FFT's.

 

[pandasaurus]

When I run stress tests at max for 20 mins in burnin test and in 3dmark firestrike extreme, temps peak 60c is while average is 52c, temps via hwinfo and yes it is stable, but my voltage is at 1.310 actually, that i did misspeak on.

 

[Darkbreeze]

Read what I posted, especially in the quoted text box. And don't just read it, READ it.


Those programs you used are NOT satisfactory for stress OR thermal testing. Use Prime95 otherwise all you are doing is giving yourself a false sense of security because they do NOT supply a steady state workload. Only those utilities I listed do that, and Prime95 version 26.6 is the best of them and (and ONLY version 26.6), should be used for that regardless what you might read on other forums. It's amazing how fast misinformation can become "truth" just because hundreds of idiots are doing something or believe in it.


I hear guys say all the time that they've run Intel burn test or some other crap, and their system is 100% stable. Then 1/2 hour into Prime95, nope, error out. If you can't run Prime95 version 26.6 for at least 15 hours without the program stopping due to errors, your CPU overclock is NOT stable regardless of what ANYBODY might tell you. They are wrong, and I will tell them to their face they are wrong.

Also, if you cannot run Prime95 v26.6 for 15 minutes without your CPU temperature exceeding 80°C, then you need to either reduce voltage or reduce multiplier. Since your voltage is already so low, if you exceed the thermal limit, then I'd drop my OC to 4.7Ghz and try again. And so on until you stay within the thermal envelope. If you get down to 4.5Ghz and are still exceeding the thermal limit, you might need to start over from the beginning, meaning, stock clocks and work your way up, until you figure out where the wall is for your CPU.

 

[pandasaurus]

Prime95 28.5 bsod’s immediately while 26.6 runs the entire test with present settings at 4.7 @1.32v cpu temp is like 65-70c, which I’m pretty sure swapping to the noctuas that should drop.

 

[pandasaurus]

I’ll have to test some more tomorrow, I need to do some more research on what you’ve posted, but as it sits right now, p95 26.6 says I’m stable at 4.75ghz.. like I said though, I’ll have to do more extensive testing tomorrow than just 2x p95 tests (well second ones almost done)

 

[pandasaurus]

Thanks again for all the info, I’ve always been told that burnintest and 3dmark were sufficient, nice to know they’re not.. I know aida64 is a joke, butyeah. Butyeah I’ll mess with it more tomorrow and see if I can stabilize at 4.8, right now 4.75 is the best I can do at 1.320v..
Since you seem to know what your talking about here, do you know much about overclocking gpu’s? Running an ASUS 1080Ti strix oc edition, overclocking in GPUTweakII, wondering what are average core and memory clock speeds? Just wondering what the average is so see if I’m in the correct ballpark on my settings

 

[Darkbreeze]

I'm not sure WHAT test you are running, or how you are running them, but there IS no "entire test". Prime will run forever pretty much if you let it. I know guys that have run the test for a week, just to make as sure as is realistically possible that the current configuration is entirely stable and no micro-errors are being introduced into the system on systems where they do advanced research and mathematics.

If you have not run Prime 26.6 for ten hours minimum, you can't even remotely say an overclock is stable. I recommend at least fifteen hours and MOST sources expect 24hrs to ensure stability and no micro-errors for general application use.

15 minutes for thermal compliance testing. If it doesn't overheat within 15 minutes, it's highly doubtful it's ever going to on Small FFT. ONLY use the Small FFT option. Any other option in Prime, including Blend, is a waste of time and is not going to purely test the CPU with a steady state workload. You can take my word for it, or you can ask Computronix, who wrote the Intel temperature guide, if that makes you feel better.

http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html

 

[pandasaurus]

I meant running 2x for 15-20minutes, was just to test the temps for now, meant I’ll have to stability test more tomorrow but for now temps are alright I guess.

 

[Darkbreeze]

Yes, if you ran Small FFT for 15 minutes and you were below 80°C the whole time then thermally, you are fine. I don't think I'd even worry about adding more/different fans unless you end up having to increase your CPU voltage to make it stable and it changes your ability to stay within the thermal envelope.

If it's not stable, and you add voltage (I recommend increasing in the smallest increments the bios will allow you to increase on each try) then when you do get it stable you need to re-test thermal compliance because very small voltage changes can have major effects on thermals.

 

[pandasaurus]

So overclocked at 4.7 with 1.31v I was able to run p95 test for 16hours, temps peak above 70 for about few a hand full of times over the 16hours but otherwise around mid ish 60s... this is running the kraken x62 with noctua ippc 3k fans in push pull as intake on my nzxt s340 elite tower; also fans in balance mode, think they top percentage reaches was only around 65%, can post hwinfo files if you’d like to take a peak.

 

[Darkbreeze]

So, they didn't drop temperatures down overall, but, the fans aren't running at full speed either. Did sound levels decrease?

Other than that, you can bet that if there are ever peaks, and those fans spin up a bit faster, they'll be able to easily handle the increased TDP.

I guess the biggest question is whether you are happy with them? You did not use the 7v reduction LNA adapters that came with them did you? They look like extensions but they're actually voltage reducers.