[SOLVED] The effectiveness of Noctua IPPC 3000s

This is a statement more than an actual question. I figured I'd post this, and others might find it useful.
Curiosity had gotten the better of me, and I wanted to see how much cooler a pc could potentially run if all the case and cpu cooler fans were replaced with IPPC 3000 pwms. I know there are plenty others curious about these monsters!
Tested in a Cooler Master H500P Mesh. My normal setup:
Intake(front): 3x NF-A14s
Intake(top): 2x NF-A14s@50%, as full speed screws with the air flow, leading to cpu running like 1-2C warmer.
^ I tried a few different fan setups, and found I got better cpu temps -VERY SLIGHT, but cooler nonetheless - over the traditional one that instead has rear and top-rear exhaust.
Exhaust(rear): 1x NF-A14
Cpu cooler: NH-D15s dual fan mode with both high speed NF-A15s
7820x(8c/16t)@ 4.3ghz, 1080ti, Corsair Vengeance LPX ram, and a 750w Seasonic Prime Titanium

For the IPPC setup, all the A14s were replaced by the 3000 version, and the A15s by the 120mm NF-F12 3000. Posting the temps of each core, I end up with:
Normal: 73, 78, 76, 84, 74, 78, 83, 74
IPPC 3000: 69, 75, 72, 80 71 74, 78 70
Gpu stayed around 55C in both cases.
Tested with Asus Realbench for 30mins. Ambient temps at 24C. I didn't win the SL with my chip. Cores 4 and 7 make it impossible to hit 4.5ghz all core on air cooling. I CAN do 4.4 though.

Also taking into account the noise levels(max speed) of the fans individually, the A14 and HS A15s are 24.6 dB/A, the A14 3000 at 41.3, and F12 at 43.5; a whole set is going to be a little louder.
With every +10 levels dB/A being ~2x louder than the previous, the 3000 IPPCs are about 4x louder(max speed, again) than their mainstream versions.


TL;DR: In my particular case(results may vary between cases), with those IPPC 3000s, my cpu ran ~4C cooler - on avg, and between the hottest core - at the cost of the case being 4x as loud :eek:
Whether that's worth it depends on the individual, but there is ONE caveat - ok, two actually: they will only get to be that loud when stress testing OCs, which users are usually(?) not around for, and if one just sets the fan curve to not go over 50-60%(about as quiet as the regulars' maximum at this level), but that defeats the purpose of getting them in the first place.
 

Darkbreeze

Titan
Moderator
For the sake of curiosity, I'd be interested in seeing you do this with a different fan configuration. I understand, theoretically, why you flew in the face of standard orientation on the case fans, but I've tested that way too and while I have seen conflicting good and bad results that way, there are thermal considerations elsewhere ESPECIALLY when overclocking, but even when not.

Ditch the 120mm fan, and put only the iPPC A14's on the CPU cooler. Remove one of the top fans, resulting in 3 front intake 3000rpm fans, one rear exhaust iPPC and one top rear exhaust iPPC. I have found, myself, that two or three top exhaust fans hamper the flow of ambient air to the CPU cooler and that I get better results with only a rear and a top rear exhaust fan, regardless of how many front or bottom intake fans are used.

I have an all NF-A14 configuration, with 6 of them being Chromax black versions of the A14 and one iPPC 2000rpm A14 in the rear of my NH-U14S and a standard A14 on the front of it. In this way, the rear pull fan can be configured to have about a 300rpm higher curve than the front fan, so that it does not become an increase on static pressure resistance to the front fans efforts to blow through the cooler, and in fact helps to reduce the static pressure resistance and pull out any dead spots.

My core temps on my 6700k@4.6Ghz are about in line with your 3000rpm results, although granted it is a much fewer core count model. Your results are good, although I'm interested in why you have a higher than ten degree variance between the lowest and highest core temp. Intel spec loosely advocates ten degree variance max to the best of my knowledge. Maybe a bad paste job. Maybe a bad mount. Maybe a bad INTERNAL TIM job, which is the most likely.
 
For the sake of curiosity, I'd be interested in seeing you do this with a different fan configuration. I understand, theoretically, why you flew in the face of standard orientation on the case fans, but I've tested that way too and while I have seen conflicting good and bad results that way, there are thermal considerations elsewhere ESPECIALLY when overclocking, but even when not.

Ditch the 120mm fan, and put only the iPPC A14's on the CPU cooler. Remove one of the top fans, resulting in 3 front intake 3000rpm fans, one rear exhaust iPPC and one top rear exhaust iPPC. I have found, myself, that two or three top exhaust fans hamper the flow of ambient air to the CPU cooler and that I get better results with only a rear and a top rear exhaust fan, regardless of how many front or bottom intake fans are used.

I have an all NF-A14 configuration, with 6 of them being Chromax black versions of the A14 and one iPPC 2000rpm A14 in the rear of my NH-U14S and a standard A14 on the front of it. In this way, the rear pull fan can be configured to have about a 300rpm higher curve than the front fan, so that it does not become an increase on static pressure resistance to the front fans efforts to blow through the cooler, and in fact helps to reduce the static pressure resistance and pull out any dead spots.

My core temps on my 6700k@4.6Ghz are about in line with your 3000rpm results, although granted it is a much fewer core count model. Your results are good, although I'm interested in why you have a higher than ten degree variance between the lowest and highest core temp. Intel spec loosely advocates ten degree variance max to the best of my knowledge. Maybe a bad paste job. Maybe a bad mount. Maybe a bad INTERNAL TIM job, which is the most likely.
I'll try it out, but how did you mount the A14s to the heatsink? Wires? Because the standard fan clips won't fit on those fans.

As for the core temp difference, I do believe it's down to the internal tim. I've been paying attention to how I'm mounting the cooler, going back and forth with 180 degree turns on the screws, tried a few different applications of paste - pea sized, thin layer, small dot(seems to work best)...
The higher I OC, the worse that variance becomes. As I mentioned in my first post, 4.5ghz isn't possible because of problem cores 4 and 7, which are neck and neck for the hottest core, while core 1 runs the coolest.
When I attempt to push 4.5, there's almost a 15C difference between 1 and 4!! I have the max core temp is set to 91C, and have yet to find a stable voltage at that frequency, because those 2 cores keep hitting the throttle limit.
Current 4.3 is stable at 1.085v, and I can get 4.4 at 1.13. I imagine I need 1.17-1.18v for 4.5, but unless I can get those 2 cores to run cooler, I won't know.
 
Nice case!

It looks like the A15 clips fit after all. I didn't think they would.
So I did the 3x intake, 2x exhaust like you told me, but it didn't appear to do much.
Ambient temps were 1C higher at 25.
70, 75, 73, 81, 71, 75, 80, 71
But I still got a lower average with 2x top intake(50% speed):
69, 74, 72, 81, 71, 75, 80, 70

After double checking my fan profiles in bios, I discovered that the fans aren't running within spec - it is stated on the box 'give or take 10%', so for the IPPC 3000s, that's 2700rpm, and 1350rpm for the normal/chromax ones.
This mobo has 7 fan headers, and an extension with 3 more.
The A14 ippc 3000s were hanging around 2500-2600(that's closer to 15%). The F12 3000s were actually within spec, at 2900-3000(one manages to hit 3030).
I have a set of chromax fans as well, and they were within spec, around 1400rpm.

^There is one exception to these results. The M.2 fan header on my board is especially bad. All fan models ran significantly below spec. The ippc 3000s ran ~2300, and the chromax at 1200.
So the reason your suggestion didn't seem to do so well is because of my motherboard; the fan headers - with the exception of M.2 - don't appear to be able to handle the power draw of the A14 3000s.
 

Darkbreeze

Titan
Moderator
No, those clips are actually slightly different than the A15 clips, because the A15 fans are asymmetrical while the A14 fans are symmetrical, so the distance from the holes on the A15 to the fins of the heatsink is somewhat different, requiring the clip to be somewhat different as well. I know they sent me some different clips, but what I can't remember, for sure, is whether I used them or simply used the A15 clips so I might have to eat crow on that statement since I can't remember for sure. LOL.

Might be a good time to go with something like this:

https://www.amazon.com/Phanteks-Universal-Fan-Controller-PH-PWHUB_02/dp/B07NHQRCRM
 
It's worth a shot. As it stands, I'm not getting my money's worth on these ippcs(the A14s, at least. The other fans appear to be fine.) if they're not running as advertised.
Looks like I'll need 2 of them though, assuming the fan headers are 1amp each, as an A14 3000 pulls just over 1/2 an amp by itself. Connecting those with y-splitters isn't a good idea.
I'll post an update once I receive the controllers and run some more tests.
 
For the sake of curiosity, I'd be interested in seeing you do this with a different fan configuration. I understand, theoretically, why you flew in the face of standard orientation on the case fans, but I've tested that way too and while I have seen conflicting good and bad results that way, there are thermal considerations elsewhere ESPECIALLY when overclocking, but even when not.

Ditch the 120mm fan, and put only the iPPC A14's on the CPU cooler. Remove one of the top fans, resulting in 3 front intake 3000rpm fans, one rear exhaust iPPC and one top rear exhaust iPPC. I have found, myself, that two or three top exhaust fans hamper the flow of ambient air to the CPU cooler and that I get better results with only a rear and a top rear exhaust fan, regardless of how many front or bottom intake fans are used.

I have an all NF-A14 configuration, with 6 of them being Chromax black versions of the A14 and one iPPC 2000rpm A14 in the rear of my NH-U14S and a standard A14 on the front of it. In this way, the rear pull fan can be configured to have about a 300rpm higher curve than the front fan, so that it does not become an increase on static pressure resistance to the front fans efforts to blow through the cooler, and in fact helps to reduce the static pressure resistance and pull out any dead spots.

My core temps on my 6700k@4.6Ghz are about in line with your 3000rpm results, although granted it is a much fewer core count model. Your results are good, although I'm interested in why you have a higher than ten degree variance between the lowest and highest core temp. Intel spec loosely advocates ten degree variance max to the best of my knowledge. Maybe a bad paste job. Maybe a bad mount. Maybe a bad INTERNAL TIM job, which is the most likely.
UPDATE, with a couple of differences
1)Reapplication of thermal paste. The previous application wasn't bad at all - a nice dollar sized spread. But I went with even less this time.
2)Corsair Vengeance LPX(2666) ram was replaced with G.Skill Ripjaws V(3466) - an exta 11mm of height. But for testing consistency, I clocked the speed and timings down matching the older kit.

3In, 2Ex:
70, 76, 73, 82, 72, 75, 81, 71
5In, 1Ex:
70, 76, 73, 82, 71, 75, 81, 72
Now, It looks like they're about the same, but that's not all. Temps have gone up slightly overall compared to the previous runs. I guess it's due to me using less paste this time around.

I also found out what was up with the low fan rpm issue. The acrylic top panel is too restrictive. Any fans I mount up top as intake are ok, but switch to exhaust, and they suffer quite a bit, running over 5% under the rated specs.
This, I believe, is why doing top exhaust isn't pulling ahead like it normally does. Not the case if I take the top off, but I don't want to run the PC that way.
To add to the issue, the small mesh strips in the top panel are FILTERED. That doesn't help, nor does it benefit exhaust fans. Perhaps it was Cooler Master's intention for the top to be intake?
 
Darkbreeze, do you, or anyone you know, have dealt with Silicon Lottery before? I'm considering sending my cpu off to them for a delid.
I plan to keep this chip for awhile, or until the next big breakthrough. I've read and watched a number of videos on the subject, but I'm still not comfortable with doing it myself.
 
Darkbreeze, do you, or anyone you know, have dealt with Silicon Lottery before? I'm considering sending my cpu off to them for a delid.
I plan to keep this chip for awhile, or until the next big breakthrough. I've read and watched a number of videos on the subject, but I'm still not comfortable with doing it myself.
I bought a delidded I5-8600K from silicon lottery.
They seem reputable, and the delidding job has given me no problems.
I think having an experienced outfit that used top quality materials is the way to go for delidding.
And, I think it gets tested and binned too.
 
I bought a delidded I5-8600K from silicon lottery.
They seem reputable, and the delidding job has given me no problems.
I think having an experienced outfit that used top quality materials is the way to go for delidding.
And, I think it gets tested and binned too.
Yeah, I've seen that option on their website, although, it DOES take away the thrill of doing it(binning) yourself.
I've seen differing discussions about how often the liquid metal needs to be replaced, but if done correctly, which is apparently difficult, it shouldn't need to be replaced very often - a one and done deal lasting until the cpu becomes obsolete.
But I don't plan on replacing it anytime soon, so I shouldn't have to worry about it?
 
Yeah, I've seen that option on their website, although, it DOES take away the thrill of doing it(binning) yourself.
I've seen differing discussions about how often the liquid metal needs to be replaced, but if done correctly, which is apparently difficult, it shouldn't need to be replaced very often - a one and done deal lasting until the cpu becomes obsolete.
But I don't plan on replacing it anytime soon, so I shouldn't have to worry about it?
I am not worried about it.
I think longevity depends on the quality of the liquid metal product used.
That might be an appropriate question to ask Silicon lottery.
Their user forum might also have that discussion.
 

Darkbreeze

Titan
Moderator
For me the deciding factor would be, if you do it yourself and ruin the CPU, too bad. If THEY ruin it, they replace it. It costs less to have the service done than to replace the CPU so it has value IMO. True, there is satisfaction in doing it yourself AND if you buy the kit or tools needed to delid, you can then do others in the future, but how often will you actually delid a CPU from any given generation?

If the answer is often, then again it has value. If you would only ever do those which are your own, then it's probably not worth it compared to the one time cost of having a reputable company do it.

For the filtration on top of the case, I'd remove it. There is no need, ever, for filtration on exhaust fans. If you use the fans as intake in the top which I think everybody here knows how I feel about that, then yes the filtration is recommended because intake filtration is universally a good thing but mostly only on systems with equal or negative pressure. If you have a ton of positive pressure which it sounds like you do with front and top intake, it might not even be necessary because of course it will naturally suppress dust intake to some degree but it sure can't hurt to have an added layer of protection against it.

If that were my case, I'd do a mod mesh modification to both the front and top panels. GamersNexus did just the front and saw about a 12 degree difference according to this.

https://www.gamersnexus.net/guides/3085-project-haf-ify-mesh-cooler-master-h500p-case-mod
 

Darkbreeze

Titan
Moderator
We had other liquid metal application results that didn’t make it to the charts: The tests above are from our third of three attempts at re-application. It can sometimes take a few tries to really get a good application of liquid metal, primarily due to surface tension and quantity of liquid metal used. We wanted the most level and stable per-core temperatures, and so took the best of each application.


The point of detailing this is that, of three new applications, we had two results that we threw away. When conducting the original test, we did the same thing: We had three applications, and then the third was the best one. This shows that there’s more variance in application of liquid metal – a whole lot more – than there is measurable difference in liquid metal aging on our platform (again, based on a case study – please note that this is not 100% conclusive for all LMs on all platforms). You should be more concerned about how well the liquid metal is applied than about its aging. That said, monitor thermals every now and then just to make sure all is well. If you’ve been running liquid metal in a system for a while, please let us know how it has aged.


There very likely is some sort of aging, but we were not able to measure it in this testing. It could be at the two-year mark, it could be different liquid metals, it could be the testing, but we did not see any meaningful change after one year of performance testing. There is far greater impact from changing the method of application, which will happen every time LM is re-applied. Most likely, if re-applying liquid metal, simple experience of application will cause more improvement than the change itself.
https://www.gamersnexus.net/guides/3359-liquid-metal-aging-one-year-test-how-often-to-replace-liquid-metal
 

Karadjgne

Titan
Herald
So far, you've basically been concentrating on cpu temps as such, and that for only 30 mins, give or take. What I'd be interested to see would be a combined effort, like in heavy gaming situations, where gpu is dumping heat for most of 1-2 hours.
 
I am not worried about it.
I think longevity depends on the quality of the liquid metal product used.
That might be an appropriate question to ask Silicon lottery.
Their user forum might also have that discussion.
Done. Waiting on a reply from them about it. Afterwards, I'll probably ship it out to 'em Tuesday or Wednesday.

For me the deciding factor would be, if you do it yourself and ruin the CPU, too bad. If THEY ruin it, they replace it. It costs less to have the service done than to replace the CPU so it has value IMO. True, there is satisfaction in doing it yourself AND if you buy the kit or tools needed to delid, you can then do others in the future, but how often will you actually delid a CPU from any given generation?

If the answer is often, then again it has value. If you would only ever do those which are your own, then it's probably not worth it compared to the one time cost of having a reputable company do it.

For the filtration on top of the case, I'd remove it. There is no need, ever, for filtration on exhaust fans. If you use the fans as intake in the top which I think everybody here knows how I feel about that, then yes the filtration is recommended because intake filtration is universally a good thing but mostly only on systems with equal or negative pressure. If you have a ton of positive pressure which it sounds like you do with front and top intake, it might not even be necessary because of course it will naturally suppress dust intake to some degree but it sure can't hurt to have an added layer of protection against it.

If that were my case, I'd do a mod mesh modification to both the front and top panels. GamersNexus did just the front and saw about a 12 degree difference according to this.

https://www.gamersnexus.net/guides/3085-project-haf-ify-mesh-cooler-master-h500p-case-mod
It's not something I'd do often. If I had a 9900k instead, I wouldn't even bother considering this, as the gains aren't as great compared to skylake-x at least.
As for the case, it's CM's H500P Mesh. I bought one with the help of this video: Awards Show: Best & Worst PC Cases of 2018. Too bad they didn't do the same thing with the top panel. So I'll look into removing that stupid acrylic and placing a mesh frame instead. Any recommendations for DIY mesh filters?

So far, you've basically been concentrating on cpu temps as such, and that for only 30 mins, give or take. What I'd be interested to see would be a combined effort, like in heavy gaming situations, where gpu is dumping heat for most of 1-2 hours.
That's due to the cpu thermals being the bigger issue in this build. 4.3ghz with no cache or gpu OC, is as far as I can take it. I can actually do 4.4, but once cache and gpu OCs have been factored in, my problem cores 4 and 7 hit the 91C temp limit I've set in bios. Once I've gotten the cpu taken care of, I definitely will be working with the gpu next - I haven't forgotten about it. Currently have a 3-step fan profile of 30C = 40%, 50C = 70%, and 70C = 100% on Afterburner at the moment. When running Realbench(main testing program), no additional voltage/120%(max) power limit, temps stay around 55C. When I add my +100 core/+150 memory OC(not a very big one, I know), it still manages to stay under 60C.
 

Karadjgne

Titan
Herald
Gotcha. My direction of thought was less on specific temps and more on patterns. Fans don't have a huge area of draw, it's a relatively weak low pressure area in front, so tend to pull from the closest available higher pressure area. Which has a tendency on modern cases to be the top rear fan port area. By setting a fan there as intake, you basically supply the exhaust fan with everything it needs, while simultaneously returning any rising heat from the gpu back on itself. That's the theory anyways, and seems to make sense. By concentrating on cpu, sure the adjacent top fan is going to supply the heatsink fan with plenty of direct air, but it's the lower half spent cooking in its own juices where my question lies. Just how much will heavy gaming and constant recycling of warmer air over extended periods of time affect the cooling vrs a standard in-out airflow pattern.
 

Darkbreeze

Titan
Moderator
Gotcha. My direction of thought was less on specific temps and more on patterns. Fans don't have a huge area of draw, it's a relatively weak low pressure area in front, so tend to pull from the closest available higher pressure area. Which has a tendency on modern cases to be the top rear fan port area. By setting a fan there as intake, you basically supply the exhaust fan with everything it needs, while simultaneously returning any rising heat from the gpu back on itself. That's the theory anyways, and seems to make sense. By concentrating on cpu, sure the adjacent top fan is going to supply the heatsink fan with plenty of direct air, but it's the lower half spent cooking in its own juices where my question lies. Just how much will heavy gaming and constant recycling of warmer air over extended periods of time affect the cooling vrs a standard in-out airflow pattern.
This may be relevant. Ran Realbench stress test overnight(8hr), ambient of 24C.
Front intake: 3x IPPC 3000s
Top intake: 3x A14 Chromax
Rear: 1x IPPC 3000
70, 77, 72, 81, 71, 74, 81, 71, and gpu was 54C avg, max 55C.
No noticeable difference from my previous tests. The recycling of warm air doesn't appear to be happening. I believe it would probably be due to most users' setups having weak pressure fans in front. I thought to reinstall the stock fans in the front, but realized that would be redundant, as due to how weak they are(airflow and pressure), temps would be higher regardless. These IPPC 3000s are just bananas!

Online, I've only ever bought it from MNPC tech, but I'm sure you can find it cheaper locally in most cases. I usually buy pieces of 12x24x.063 round perforated aluminum or steel.

Not particularly cheap, but again, you can likely find it cheaper by looking around.

https://mnpctech.com/case-mods-gaming-pc-liquid-modding-custom-computer-mnpctech-overclock-cooling-fan-grills/modders-mesh-perforated-aluminum-steel-honeycomb-grill-intake-front-fan-slotted-diamond-round-hex/modders-mesh.html
I'm going to check the nearby Lowe's for one. I see that I can actually move these holders on the underside of the top panel and slide the plexiglass out. Then I can slide a mesh panel in.
 

Darkbreeze

Titan
Moderator
They have perforated metal at lowes, where I can be found several times per day (But what your local store has I can't say) but I've never seen the right kind there.

They have this, but it's only about a 20% open pattern. Not good enough.

https://www.lowes.com/pd/M-D-24-in-x-3-ft-Aluminum-Sheet-Metal/1000243337

You really want something with 50% or higher open area like some of the patterns seen here:

http://www.accuratealloys.com/round-hole-perforated-metal.html

Local fabrication, HVAC/sheet metal installers or metal specific yards will be more likely to have similar products and will probably be significantly cheaper than Lowe's or Home depot.
 
They have perforated metal at lowes, where I can be found several times per day (But what your local store has I can't say) but I've never seen the right kind there.

They have this, but it's only about a 20% open pattern. Not good enough.

https://www.lowes.com/pd/M-D-24-in-x-3-ft-Aluminum-Sheet-Metal/1000243337

You really want something with 50% or higher open area like some of the patterns seen here:

http://www.accuratealloys.com/round-hole-perforated-metal.html

Local fabrication, HVAC/sheet metal installers or metal specific yards will be more likely to have similar products and will probably be significantly cheaper than Lowe's or Home depot.
Just got back from there, and saw that exact same sheet. There were also a couple others with some fancy patterns, but no better than that first one. Some door/window screens, bug screens, etc., not what I needed.
I'm just gonna make it easier for myself and order the sheet from mnpctech. I can't go and drive wherever I want at the moment - doctor's orders... have to get a relative to do most of the driving for me.
I've already got a dremel and carbo-titanium scissors at home, assuming that will be enough to cut and trim, if need be.

EDIT: Some one with SL finally got back with me and told me that the LM shouldn't need to be reapplied, but if something does happen, I can have them do it again... and they've apparently only had a couple of reports of 'potential' thermal degradation in the last five years, for what it's worth.
So I'm sending my chip off tomorrow morning.
 
Last edited:

Darkbreeze

Titan
Moderator
Cool man. I understand not being able to drive. I've had two shoulder surgeries, to different shoulders, and a biceps tenodesis surgery, in the the last four years. Let us know how it goes and if you have any questions that I CAN answer, of course I'm happy to. Modifications are fun. Probably my favorite part of being an enthusiast. I've done quite a few of them over the years.
 
Reactions: Phaaze88
UPDATE: Got the SL delidded cpu from the mail this morning, and gave it a spin. Ambient temps were 25C. Setup was the same as the last:
3x ippc 3000 front intake
3x A14 Chromax top intake(top exhaust doesn't, and will not do any better than this until I get the plexiglass replaced from the case's top panel)
1x ippc 3000 rear
63, 63, 67, 69, 66, 67, 67, 64
I gained about -10C average overall. As for the quality of the chip, it wasn't as impressive compared to SL's recorded binning of these cpus.
4.5@1.225 = 100%(% refers to the number capable among their records)
4.6@1.237 = 98%
4.7@1.250 = 76%
4.8@1.262 = 41%
4.9@1.275 = 12%
Mine landed at 4.6. Shucks!
I still have the option to try and OC each core past that - though I don't see much point in doing so after the binning results - and there's still the cache and memory to tweak as well.
Now I need to wait for the mesh I ordered to come in, and I can get started on that as well.
 

ASK THE COMMUNITY

TRENDING THREADS

Latest posts