Cooling Noctua NH-P1 passive cooler. To have or not to have?

[tinyitx]

I am getting more interested in this newly released Noctua fanless cooler, NH-P1.

I do understand that this monstrosity will not be popular here or in SFF circle as it does not fit into many SFF cases.
Nonetheless, I wish to see what you guys think. Both pros and cons opinions are welcome.

Here are 2 reviews that I have read, so far.

Noctua NH-P1 Passive CPU Cooler Review

We at eTeknix review the newly released Noctua NH-P1. It’s long anticipated passive (fanless) high-end CPU cooling solution!

www.eteknix.com

(this one has a Youtube video review as well as a written review and they are not identical)

 

[REVOCCASES]

I am getting more interested in this newly released Noctua fanless cooler, NH-P1.

I do understand that this monstrosity will not be popular here or in SFF circle as it does not fit into many SFF cases.
Nonetheless, I wish to see what you guys think. Both pros and cons opinions are welcome.

Here are 2 reviews that I have read, so far.

Noctua NH-P1 Passive CPU Cooler Review

We at eTeknix review the newly released Noctua NH-P1. It’s long anticipated passive (fanless) high-end CPU cooling solution!

www.eteknix.com

(this one has a Youtube video review as well as a written review and they are not identical)

I am actually very tempted to get one of those. I mean watercooling is fun but... just look at that cooler...

How do you think if I make an NH-P1 optimized version of my BIG2 Case?

It wouldn't be the smallest case with around <=18L but I think the overall layout could achieve pretty good and (semi-) silent results with the NH-P1 (plus one of those big 3 slot GPUs and one SFX-L PSU).

I just wonder what would work better: making it some kind of open-bench-case-hybrid with even more vent holes |OR| going for solid panels left and right with air intake from front and bottom and forced exhaust from the top...

 

[tinyitx]

How do you think if I make an NH-P1 optimized version of my BIG2 Case?

It wouldn't be the smallest case with around <=18L but I think the overall layout could achieve pretty good and (semi-) silent results with the NH-P1 (plus one of those big 3 slot GPUs and one SFX-L PSU).

I think it would be good if the case has a clear air path of 'bottom-in and top-out' to match with a natural convection pattern inside the case.
Additionally, it might work better if the motherboard is moved to the top. With the mobo and NH-P1 at the top, a natural convection pattern can happen locally. I have made a quick sketch to show my idea.

I just wonder what would work better: making it some kind of open-bench-case-hybrid with even more vent holes |OR| going for solid panels left and right with air intake from front and bottom and forced exhaust from the top...

Given your swiftness to generate a working prototype, why not make one of each and then test them to find out?

 

[REVOCCASES]

I think it would be good if the case has a clear air path of 'bottom-in and top-out' to match with a natural convection pattern inside the case.
Additionally, it might work better if the motherboard is moved to the top. With the mobo and NH-P1 at the top, a natural convection pattern can happen locally. I have made a quick sketch to show my idea.


Given your swiftness to generate a working prototype, why not make one of each and then test them to find out?

That's an very interesting layout idea!

But doesn't Noctua and that review you posted say that bottom to top airflow is overall preferred?

Do you know if the cooler could be mounted with the fins aligned correctly if I have the Motherboard I/O at the bottom like in the BIG2?

 

[tinyitx]

Yes, bottom to top airflow is preferred (ie, when the entire case airflow is considered).

My sketch shows a layout with the display card's IO pointing down. So, the cabling there might not allow a clean and neat design of a bottom 180/200mm intake fan. But, if the case is tall enough, I suppose such a bottom intake fan can still be incorporated.

And, a 'yes' to your last question. Actually, it depends on the motherboard itself. The cooler can be installed in standard Noctua's ''Orientation A'' or ''Orientation B".

In the video, 3:52 shows the cooler installed with its plane of fins pointing in the direction of the mobo IO. This, of course, is not optimal and not preferred as natural convection is restricted. This cooler orientation is the 'standard' or default setup for this cooler though (since the mobo IO is not pointing down for most cases).

In the same video , 5:10 shows the cooler with its plane of fins perpendicular to the mobo IO. This is what you would like in the BIG2 where the mobo IO points downward. Here, natural convection happens with an direct vertical path.

 

[REVOCCASES]

I did some Prime95 testing in the BIG2 with my R5 2600 and Thermalright SilverArrow in passive- (all fans OFF) and semi- passive mode (fans set to lowest possible RPM).

Although the SilverArrow is not meant to be a passive cooler it actually performs pretty good.

Looking at Noctuas NSPR and CPU compatibility chart for the P1 I'm not really sure if it's worth for me to get one or make a case for it. With the SilverArrow I was able to keep the CPU well below 75C in passive mode without boost and same goes for semi-passive mode with enough headroom for boost.

Although I really like the design & look of the P1 - overall I think a cooler like the NH-D15 or SilverArrow should be a better value than the P1. They already come with a fan in the box at a lower price point and combined with a low wattage CPU it seems you can also run them (semi-)passive if you like to. Plus, if you upgrade to a higher wattage CPU you don't have to get a new cooler but only need to adjust your fan curve or add (more) fans to the cooler.

 

[tinyitx]

Well, I suppose the R5 2600 is a 'low wattage' CPU with its 65W TDP. TDP figures might not be very meaningful for this purpose but I still use it anyway, just for the sake of having a relative number to use.

Noctua designs the fanless NH-P1 to handle 120W TDP CPU in a fanless case. Reviews show that this figure is quite accurate. I plan is to cool a 9900K, at stock setting first and then perhaps revert to its slightly overclocked speed. Intel says 9900K has PL1 95W and PL2 119W. So, NH-P1 should cool well for my plan.

Additionally, I am using an open frame (ZS OC v1). This provides an ideal environment as natural convection happens without restriction. Even with no case fan, hot air rises up optimally.

My system is not fanless though, even with a fanless NH-P1 as my SF600 has a fan and the MSI 2070 Gaming Z has twin fans. But, both are semi-passive as both have zero rpm mode enabled. These fans will only spin when the system is under heavy load like gaming. So, with a fanless NH-P1, my system is 100% zero rpm until gaming starts.

There is a Chinese saying ''我不入地狱谁入地狱''. Literal translation is ''Who would go to hell if not me''. It means ''if someone has to take the tough stuff, let me be the one''. With this spirit, I will get one to test if it can cool my 9900K well. Stay tuned, folks.

 

[REVOCCASES]

Well, I suppose the R5 2600 is a 'low wattage' CPU with its 65W TDP. TDP figures might not be very meaningful for this purpose but I still use it anyway, just for the sake of having a relative number to use.

Noctua designs the fanless NH-P1 to handle 120W TDP CPU in a fanless case. Reviews show that this figure is quite accurate. I plan is to cool a 9900K, at stock setting first and then perhaps revert to its slightly overclocked speed. Intel says 9900K has PL1 95W and PL2 119W. So, NH-P1 should cool well for my plan.

Additionally, I am using an open frame (ZS OC v1). This provides an ideal environment as natural convection happens without restriction. Even with no case fan, hot air rises up optimally.

My system is not fanless though, even with a fanless NH-P1 as my SF600 has a fan and the MSI 2070 Gaming Z has twin fans. But, both are semi-passive as both have zero rpm mode enabled. These fans will only spin when the system is under heavy load like gaming. So, with a fanless NH-P1, my system is 100% zero rpm until gaming starts.

There is a Chinese saying ''我不入地狱谁入地狱''. Literal translation is ''Who would go to hell if not me''. It means ''if someone has to take the tough stuff, let me be the one''. With this spirit, I will get one to test if it can cool my 9900K well. Stay tuned, folks.

True, TDP doesn't mean much these days - even Thermalright doesn't mention it on their new coolers any more and instead saying "thermal power capacity".

Guess I should have mentioned that my CPU would draw between 70 to 85W during my tests.

Looking forward to your review of the P1!

 

[tinyitx]

While waiting, may watch these 4 more reviews:-
a. cooling 9900K

b. cooling 11700 (written in Chinese, but the tables and graphs and pics are universally understood)

[開箱] 貓頭鷹 NH-P1 無風扇CPU散熱器

PTT好讀版 https://www.ptt.cc/bbs/PC_Shopping/M.1624559144.A.4D7.html 網誌無廣告版 https://ohmy-bear.blogspot.com/2021/06/noctua-nh-p1-cpu.html 以下正文 貓頭鷹最新款的CPU散熱器產品，敲碗敲到破的無風扇被動式散熱器，終於姍姍來遲啦！ 二話不說，在AMAZON上架的第一天立刻給他訂購下去！

www.ptt.cc

c. cooling 3700X

d. cooling 3600 (Edit)

 

[REVOCCASES]

While waiting, may watch these 4 more reviews:-
a. cooling 9900K

b. cooling 11700 (written in Chinese, but the tables and graphs and pics are universally understood)

[開箱] 貓頭鷹 NH-P1 無風扇CPU散熱器

PTT好讀版 https://www.ptt.cc/bbs/PC_Shopping/M.1624559144.A.4D7.html 網誌無廣告版 https://ohmy-bear.blogspot.com/2021/06/noctua-nh-p1-cpu.html 以下正文 貓頭鷹最新款的CPU散熱器產品，敲碗敲到破的無風扇被動式散熱器，終於姍姍來遲啦！ 二話不說，在AMAZON上架的第一天立刻給他訂購下去！

www.ptt.cc

c. cooling 3700X

d. cooling 3600 (Edit)

Thanks for sharing. I took below screenshot from the video with the Ryzen 3600 which should be a good enough comparison to my tests with the 2600, right?

If I read this correctly it means that his 3600 reached 80C while pulling only 60W.

I will do some more testing with the SilverArrow in passive mode but - unless I'm missing something - seems it would beat the P1.

 

[thelaughingman]

@tinyitx JUST DO IT! someone else already done it

 

[tinyitx]

Thanks for sharing. I took below screenshot from the video with the Ryzen 3600 which should be a good enough comparison to my tests with the 2600, right?

If I read this correctly it means that his 3600 reached 80C while pulling only 60W.

I will do some more testing with the SilverArrow in passive mode but - unless I'm missing something - seems it would beat the P1.

Sorry, I do not quite get it what you meant. Please elaborate how the Silver Arrow would beat the P1 in passive mode.

@tinyitx JUST DO IT! someone else already done it

His P1 is oriented in the worst possible way out of 4, with the fins in a horizontal position (thus, natural convection is hindered as hot air from the first fin down does not have a clear straight upward path of movement....see the Fully Silent PCs Youtube video for demonstration....I wonder why he did it that way. Maybe he could not rotate 90-degree to fit the P1 in?!) and the heat pipes pointing down (ie lower than the heatsource...but this factor should not be as dominate as the P1 installation orientation).

I wonder what his fanless display card is? My 2070 is the least powerful card that I am happy with. I do not want to downgrade any lower.
And I do not think I can go totally fanless as a (Seasonic) 500W should not be able to handle my 9900K + 2070 very comfortably.
But to his big big credit, that guy has done it, totally fanless! I suppose this is his main point.

Now, this guy could do it with a 5600X (TDP 65W). Can I do it with a 9900K (PL1 95W, PL2 119W)?
BTW, this bad boy has just sneaked into my home silently.

 

[rfarmer]

Sorry, I do not quite get it what you meant. Please elaborate how the Silver Arrow would beat the P1 in passive mode.



His P1 is oriented in the worst possible way out of 4, with the fins in a horizontal position (thus, natural convection is hindered as hot air from the first fin down does not have a clear straight upward path of movement....see the Fully Silent PCs Youtube video for demonstration....I wonder why he did it that way. Maybe he could not rotate 90-degree to fit the P1 in?!) and the heat pipes pointing down (ie lower than the heatsource...but this factor should not be as dominate as the P1 installation orientation).

I wonder what his fanless display card is? My 2070 is the least powerful card that I am happy with. I do not want to downgrade any lower.
And I do not think I can go totally fanless as a (Silverstone) 500W should not be able to handle my 9900K + 2070 very comfortably.
But to his big big credit, that guy has done it, totally fanless! I suppose this is his main point.

Now, this guy could do it with a 5600X (TDP 65W). Can I do it with a 9900K (PL1 95W, PL2 119W)?
BTW, this bad boy has just sneaked into my home silently.

What are your thoughts on his PSU location, won't the heat from the PSU go straight up through the heatsink?

 

[REVOCCASES]

Sorry, I do not quite get it what you meant. Please elaborate how the Silver Arrow would beat the P1 in passive mode

With the SilverArrow I could pull 70W without hitting 80C (ambient temp 28C) in passive mode.

The HWINFO screenshot from the P1 video shows 80C while pulling only 60W.

So I mean it looks to me that with the SilverArrow I could pull more wattage while keeping my CPU cooler. But I know, it's not the same test conditions and the SilverArrow would probably perform much worse in another orientation... Still interesting IMO, going to do some more testing...

 

[tinyitx]

What are your thoughts on his PSU location, won't the heat from the PSU go straight up through the heatsink?

Seasonic recommends that it is best to install their fanless PSU with the top cover facing UPWARD. This is logical and makes sense as a fanless PSU relies on natural convection to dissipate the heat. The perforated top cover could allow hot air to rise naturally by convection.
Of course, if the case has some case fans to provide some air movement inside the case, the PSU can be installed in such a way to take advantage of that air flow. But, I bet the focus of Seasonic is at a totally fanless situation as this will be the 'worst' case scenario for their fanless PSU to ever encounter.

Now, I take a closer look at that guy's PSU orientation and note that his installation is also the worst possible way!!
The reason is, the PSU cables are pointing up, meaning the green pcb and the solid top cable connectors side are blocking natural convection. Convection now has to take not the least path of resistance but through the side perforated panels. This is not optimal to the PSU itself.

Frankly, if his P1 is oriented optimally (ie having fins vertically), I would rather have the hot air coming up from the PSU to go through the fins of the P1. My argument is:- I would rather have hot air natural convection in an easy way (so to speak) than not to have any convection at all (or, to have natural convection is a hard way). I figure that once natural convection is formed, cool air will continuously pulled in near the bottom and air will rise upward to cool whatever components it comes into contact with.

Sorry for this 'wall of text'!

Anyway, I am glad you asked the question as now I have looked into this Seasonic fanless PSU.
I now discover, Seasonic offers 600W and 700W fanless ATX PSUs. And guess what, my ZS OC v1 open frame can accommodate ATX PSU! So, I could change my Corsar SF600 to a fanless ATX Seasonic!
But, even so, I still cannot make my 2070 fanless...lol

 

[tinyitx]

With the SilverArrow I could pull 70W without hitting 80C (ambient temp 28C) in passive mode.

The HWINFO screenshot from the P1 video shows 80C while pulling only 60W.

So I mean it looks to me that with the SilverArrow I could pull more wattage while keeping my CPU cooler. But I know, it's not the same test conditions and the SilverArrow would probably perform much worse in another orientation... Still interesting IMO, going to do some more testing...

Now I see what you meant. Has to be some testing error or a 'comparing apples to oranges' case.

But, if your do confirm your findings, you might make some money by selling your results to Thermalright so that they can advertise their SAs having beating the P1 passively
Or, you can sell your findings to Noctua. I bet they would pay you $ to keep your mouth shut. Otherwise, their years of research on the P1 and the potential profit would all vaporize...lol

 

[REVOCCASES]

Now I see what you meant. Has to be some testing error or a 'comparing apples to oranges' case.

But, if your do confirm your findings, you might make some money by selling your results to Thermalright so that they can advertise their SAs having beating the P1 passively
Or, you can sell your findings to Noctua. I bet they would pay you $ to keep your mouth shut. Otherwise, their years of research on the P1 and the potential profit would all vaporize...lol

Well, that's my point. If a big tower cooler like the SilverArrow or D15 could already be run in passive mode up to a certain, similar power draw then I probably will not get a (more expensive) P1 - despite I love the looks of that beast.

 

[tinyitx]

Well, that's my point. If a big tower cooler like the SilverArrow or D15 could already be run in passive mode up to a certain, similar power draw then I probably will not get a (more expensive) P1 - despite I love the looks of that beast.

Fair enough!
Maybe before switching to the P1 from my current NH-C14S, I would run some benchmark to test the C14S. I do not think it can run my 9900K (even at stock) totally fanless. Maybe I will try 300 rpm to see if what the CPU temp will rise up to.

Anyway, I took a closer look at the Seasonic fanless 600W and 700W ATX PSUs. Both are Titanium rated and both receive very good reviews with the major cons as their high prices. Well, I guess not many good things come in cheap. And Seasonic covers them with 12-year warranty! My SF600 has 'only' 7 years and I think Corsair covers their best units with 10. So, I suppose, if one wants to go fanless on the PSU, Seasonic should be the prime candidate.

 

[thelaughingman]

Anyway, I took a closer look at the Seasonic fanless 600W and 700W ATX PSUs. Both are Titanium rated and both receives very good reviews with the major cons as their high prices. Well, I guess not many good things come in cheap. And Seasonic covers them with 12-year warranty! My SF600 has 'only' years and I think Corsair covers their best units with 10. So, I suppose, if one wants to go fanless on the PSU, Seasonic should be the prime candidate.

just do it! on the whole Seasonic PSUs have much better reputation than Corsair in Vietnam too. Would love for Seasonic to launch an SFX(-L) flagship to compete against the SF750

 

[tinyitx]

Removing the NH-C14S was a breeze. After 2+ years, the thermal paste (Gelid GC-Extreme) still looks OK.


A pre cleaned up 9900K


A post cleaned up 9900K (with Noctua SecurFirm2+). I suppose, the main (only?) difference between 2 and 2+ is that screws are now torx-based.


NH-P1 in its full glory. So massive, so industrial looking, so monstrous, so indestructible.
Current installation is the default orientation. It clears the rear Asus Z390i integrated IO shield and the heatpipes just barely clears the RAM as well. I am debating if I would install it in the opposite way, making the 'front panel with Noctua marking' facing front. This might looks better but the RAM will be 'hidden'. Not sure if I like that.

Will do some testing later today.