UPDATE: D15 is absolutely more reliable and makes more sense for most unless you get a top of the line liquid cooled set up. A cheap liquid cooled set up is not better than the d15.
Thanks for this. Could you add another test? Just leave the wattage uncapped and see which cooler achieves the best bench results (render times etc). This way we will see the real performance of the cooler.
Working on this right now but I plan to take a more sophisticated approach and add it to the model system I currently created. Currently I have 4 benchmark applications that i plan to collect data for.
Hi, so I am in between choosing these coolers, my CPU will be 7800X3d, but the thing is during summers the ambient temperatures of the room might go to 35C. Which one will be recommended in that scenario. Will only be using it for gaming, no synthetic loads. Thanks for this short and to the point review.
My model says ambient temps impact cooler by a coefficient of .6. So that means I would add (35-22)*0.6=7.8 to the graphs on what would I expect. But that temp is way outside the range in what I consider so I can’t guarantee accuracy on that metric. My website has results on the 7800x3d if you want to see the temps on that cpu. The max ppt I show got 4.8ghz on cinebench.
I might have found an answer to a generalized version of a question alluded to in my last comment: why is 30 dBA the least measurement value of the noise from a fan, CPU cooler, etc. shown in almost all tech reviews on the Internet? It's the limitation of sound pressure level meters readily and widely available. 30 dBA is the lowest sound level measurable by such meters. Such meters show 30 dBA for any sound pressure level equal to or less than 30 dBA. Your ambient level might really be below 30 dBA. Your PSU on eco mode might be really slightly above your ambient level. A fan tested might go below 30 dBA. All these possible state of affairs cannot be detected by a SL meter that cannot measure any sound level below 30 dBA. Although you need a professional sound level meter to measure sound level below 30 dBA accurately, you can estimate sound level below 30 dBA by measuring sound level at a distance less than 1 m, which is the standard distance in sound level measurement, and calculating the sound level at 1 m by this equation. SPL2 = SPL1 - 20log10r2/r1 where SPLn is the sound level at the distance of rn m from the sound source (n=1,2) For instance, If your SPL meter shows 31 dBA at 0.5m, then its sound level is estimated to be approximately 25 dBA at 1m.
this "review" seems odd to me. Why use a previous generation mid-range CPU? It seems disingenuous to me to cherry pick one specific CPU for results and make a genuine case for cooler effectiveness. One sample point never makes a convincing argument. you need correlation between multiple points to infer a definitive result. If you called this a review of the 5800X under both liquid and air cooling, that would be a more accurate description. Both coolers are rated for much higher TDP loads than this puts out. I realize it costs money to have multiple setups for reviews, but alluding to the title, saying this is a review of the cooling effectiveness of these setups is wholly inaccurate.
Thanks for your note, a few things: • if your interested in seeing multiple CPU's under these coolers, check out my website where I have results across multiple CPU's. • Although some may like to see across multiple CPU, view time metrics show that most users only want to see a specific set of results, duplicative content that you are suggesting would not be watched by most users who come across the video. • Currently this video focuses on tail end search, putting massive effort with questionable value gain isn't a good use of time. In short, my videos try to be short sweet and to the point, while tail-end compares you are requesting are addressed by my website where its easier to propagate to much more content. For the video's compare: pcanalytics.com/product/coolers/compare/arctic/liquid-freezer-ii-360-a-rgb/noctua/nh-d15/ For available CPU's to see for this compare: pcanalytics.com/product/coolers/compare/select-cpu/arctic/liquid-freezer-ii-360-a-rgb/noctua/nh-d15/cpu-1/
I have NH-D15, but now want upgrade to i7-14700k, so I think I need better cooling or not. Thanks I found that - Arctic Liquid Freezer II 360 is better for 3-4C
This is mostly about Arctic Freezer AiO having lower fin density, than standard 38mm radiators - and that´s probably due to the fans not being very strong in terms of static pressure (most of official fan specs out there are pure kool-aid, thanks to Cybenetics we can now see, how most fans on the market are quite far from those).
Yeah, but did you install the rev 4 offset brackets for the Arctic liquid Freezer II 360 which brings the offsets down a bit more covering the CCD or CCD’S which get hot?
@@pc_analytics that’s nuts my Ryzen 9 5900x idles at 26c and never ever goes over 65c on an Arctic Liquid Freezer II 360 rev 4 with the offset brackets and Arctic MX-6 thermal paste and the cl 14 3600mhz 4x8gb overclocked idles at 23c never goes over 27c but I have a 120mm fan over it. I found my PBO2 limits PPT 195 EDC 140 TDC 130 no undervolting no scaler left on auto no CO no extra mhz ran Cinebench r23 30-minute throttle test and scored 23,567mc and in ST 1,624. All stable gaming on a rtx 4070 TI Super at 1440p.
This is not a good CPU for such test and does not shows coolers real strengths. The real contenders are i9-14900K and Ryzen 9 7950X. Why would I buy AIO water cooling for Ryzen 7 while we have Thermalright Phantom Spirit 120?
I wonder if you could use a fanless PSU or one whose fan does not spin under 50% load in your test since the noise level of a cooler at a low fan speeed may not be accurately measured as it is likely to be below the noise level from a PSU whose fan does not stop. For instance, your graph for the comparison under the idle condition often show that the noise level of two coolers compared are the same at the lower RPM, but I suspect their noise levels may sometimes be different but just not detected as they are both below the noise floor created by PSU fan's noise level. PSU's fan hides the true acoustic performance of a cooler, so to speak.
My PSU is on eco mode, which results in the fan to be below ambient levels and not picked up during my testing. My testing is done in a quiet closet where I hit about 30 DBa. I intentionally floor graphs to removed ambient variations from impacting graphs. If you are in a space below 32 dba, thats quite impressive.
@@pc_analytics Thank you for your reply and description of your test environment and procedures. I'm sorry for misunderstanding them. As I cannot find your PSU in Cybernatics's efficiency & noise level certification database, I looked into their report of similar PSUs: EVGA SuperNOVA 850 G6 and 750 GA. Based on these models, I surmise that your PSU's noise level is under 30dBA under the eco mode but might sometimes be close to 30dBA if it's more similar to 750 GA. I'm not sure what you exactly mean by "floor graphs," but it now seems that the difference between these two coolers at lower fan speeds may actually be so subtle as to be eliminated by some compensatory adjustment in creating a graph. I myself use Silverstone NJ700, a fanless PSU, NH-P1, NF-A12x25 LS-PWM, and NF-A12x25 PWM now.
After posting the above comment, it occured to me that your ambient noise level might be determined by your air-conditioner. You might need to turn it on even if you don't normally have to in order to maintain a certain ambient temperature for testing. (At this time of the year, I do not turn on an air-conditioner thanks to the climate of the area I live in.)
@@osardegozar2581 nope. I turn that bad boy off and do multiple runs and compare results over runs to ensure I’m unrelated outside noise is not impacting results.
@@pc_analytics Then, given all the above, your PSU seems to be the only source of noise I can conceive of that determines your ambient level. EVGA SuperNOVA 850 GA may perhaps be close to 750 GA. Perhaps PSU fan makes a greater contribution to the ambient level in an open-air case. Let me note that I do understand you've done everything you could with your current setup to ensure accurate measurement. I just want to make sense of noise levels of these coolers shown for a fan speed below 45%.
Noctua ram clsarance is kne challenge. Yet low profile non rgb just fine I prefer air cooler than Aio anh times Simply blow external.air cooler in front of mesb panel who also has intake fan to blow air jnti the system. By that thougb cosr a bit of power yet at the oing term I manage to remove the short curcuit incident. If the cooler air.from that air cooler is cold enougb and the wind is strong enough, even at maximum ooad no throttle
@@pc_analytics and the speed you flawless talk that down MACHINE 🤣but exactly what i needed since i'm currently diggig through hundreds of reviews building a pc
One of the only analysis videos on youtube where I'm playing back at 1x. Clear, concise, and instructive. Thank you.
Thanks for a quick & easy review! I've had the D14 & D15 for 8 years now. Don't think I'll ever go for liquid cooling.
Is the d15 still running after 8 years of use??
@@ravinkrishnan3183D15 is quiet & cool as ever. Eventually a better cooler will be released?
UPDATE: D15 is absolutely more reliable and makes more sense for most unless you get a top of the line liquid cooled set up. A cheap liquid cooled set up is not better than the d15.
I think I will also stay on Air cooler, Noctua is Cooler God :D
@@glbernini0 no need to change if its already perfect
1x speed is already fast on this video... WOW!!! thanks!
Thanks for this. Could you add another test? Just leave the wattage uncapped and see which cooler achieves the best bench results (render times etc). This way we will see the real performance of the cooler.
Working on this right now but I plan to take a more sophisticated approach and add it to the model system I currently created. Currently I have 4 benchmark applications that i plan to collect data for.
Hi, so I am in between choosing these coolers, my CPU will be 7800X3d, but the thing is during summers the ambient temperatures of the room might go to 35C. Which one will be recommended in that scenario.
Will only be using it for gaming, no synthetic loads. Thanks for this short and to the point review.
My model says ambient temps impact cooler by a coefficient of .6. So that means I would add (35-22)*0.6=7.8 to the graphs on what would I expect. But that temp is way outside the range in what I consider so I can’t guarantee accuracy on that metric. My website has results on the 7800x3d if you want to see the temps on that cpu. The max ppt I show got 4.8ghz on cinebench.
I might have found an answer to a generalized version of a question alluded to in my last comment: why is 30 dBA the least measurement value of the noise from a fan, CPU cooler, etc. shown in almost all tech reviews on the Internet?
It's the limitation of sound pressure level meters readily and widely available. 30 dBA is the lowest sound level measurable by such meters. Such meters show 30 dBA for any sound pressure level equal to or less than 30 dBA. Your ambient level might really be below 30 dBA. Your PSU on eco mode might be really slightly above your ambient level. A fan tested might go below 30 dBA. All these possible state of affairs cannot be detected by a SL meter that cannot measure any sound level below 30 dBA.
Although you need a professional sound level meter to measure sound level below 30 dBA accurately, you can estimate sound level below 30 dBA by measuring sound level at a distance less than 1 m, which is the standard distance in sound level measurement, and calculating the sound level at 1 m by this equation.
SPL2 = SPL1 - 20log10r2/r1
where SPLn is the sound level at the distance of rn m from the sound source (n=1,2)
For instance, If your SPL meter shows 31 dBA at 0.5m, then its sound level is estimated to be approximately 25 dBA at 1m.
Ah, thank you for that. I wasn’t aware of the floor.
this "review" seems odd to me. Why use a previous generation mid-range CPU? It seems disingenuous to me to cherry pick one specific CPU for results and make a genuine case for cooler effectiveness. One sample point never makes a convincing argument. you need correlation between multiple points to infer a definitive result. If you called this a review of the 5800X under both liquid and air cooling, that would be a more accurate description. Both coolers are rated for much higher TDP loads than this puts out. I realize it costs money to have multiple setups for reviews, but alluding to the title, saying this is a review of the cooling effectiveness of these setups is wholly inaccurate.
Thanks for your note, a few things:
• if your interested in seeing multiple CPU's under these coolers, check out my website where I have results across multiple CPU's.
• Although some may like to see across multiple CPU, view time metrics show that most users only want to see a specific set of results, duplicative content that you are suggesting would not be watched by most users who come across the video.
• Currently this video focuses on tail end search, putting massive effort with questionable value gain isn't a good use of time.
In short, my videos try to be short sweet and to the point, while tail-end compares you are requesting are addressed by my website where its easier to propagate to much more content.
For the video's compare:
pcanalytics.com/product/coolers/compare/arctic/liquid-freezer-ii-360-a-rgb/noctua/nh-d15/
For available CPU's to see for this compare:
pcanalytics.com/product/coolers/compare/select-cpu/arctic/liquid-freezer-ii-360-a-rgb/noctua/nh-d15/cpu-1/
Are you using the rev 4 offset brackets with the ALF II and the 7mm offset brackets with the NH-D15?
I have NH-D15, but now want upgrade to i7-14700k, so I think I need better cooling or not. Thanks
I found that - Arctic Liquid Freezer II 360 is better for 3-4C
This is mostly about Arctic Freezer AiO having lower fin density, than standard 38mm radiators - and that´s probably due to the fans not being very strong in terms of static pressure (most of official fan specs out there are pure kool-aid, thanks to Cybenetics we can now see, how most fans on the market are quite far from those).
Yeah, but did you install the rev 4 offset brackets for the Arctic liquid Freezer II 360 which brings the offsets down a bit more covering the CCD or CCD’S which get hot?
I used the brackets that came out of the box.
@@pc_analytics that’s nuts my Ryzen 9 5900x idles at 26c and never ever goes over 65c on an Arctic Liquid Freezer II 360 rev 4 with the offset brackets and Arctic MX-6 thermal paste and the cl 14 3600mhz 4x8gb overclocked idles at 23c never goes over 27c but I have a 120mm fan over it. I found my PBO2 limits PPT 195 EDC 140 TDC 130 no undervolting no scaler left on auto no CO no extra mhz ran Cinebench r23 30-minute throttle test and scored 23,567mc and in ST 1,624. All stable gaming on a rtx 4070 TI Super at 1440p.
@@sporadicentertainment9253 thanks for sharing. it is a great cooler
This is not a good CPU for such test and does not shows coolers real strengths. The real contenders are i9-14900K and Ryzen 9 7950X. Why would I buy AIO water cooling for Ryzen 7 while we have Thermalright Phantom Spirit 120?
I have the Noctua d15 for 13700k but will Switch to a freesee 2
this vid was legit interesting
So, there's no performance reason to go for an AIO?
Other than the look or space
@@pc_analytics what about gamernexus reviews? i thought a fan cooler would be enough for me but their tests show better results for aio?
I wonder if you could use a fanless PSU or one whose fan does not spin under 50% load in your test since the noise level of a cooler at a low fan speeed may not be accurately measured as it is likely to be below the noise level from a PSU whose fan does not stop. For instance, your graph for the comparison under the idle condition often show that the noise level of two coolers compared are the same at the lower RPM, but I suspect their noise levels may sometimes be different but just not detected as they are both below the noise floor created by PSU fan's noise level. PSU's fan hides the true acoustic performance of a cooler, so to speak.
My PSU is on eco mode, which results in the fan to be below ambient levels and not picked up during my testing.
My testing is done in a quiet closet where I hit about 30 DBa. I intentionally floor graphs to removed ambient variations from impacting graphs. If you are in a space below 32 dba, thats quite impressive.
@@pc_analytics Thank you for your reply and description of your test environment and procedures. I'm sorry for misunderstanding them. As I cannot find your PSU in Cybernatics's efficiency & noise level certification database, I looked into their report of similar PSUs: EVGA SuperNOVA 850 G6 and 750 GA. Based on these models, I surmise that your PSU's noise level is under 30dBA under the eco mode but might sometimes be close to 30dBA if it's more similar to 750 GA. I'm not sure what you exactly mean by "floor graphs," but it now seems that the difference between these two coolers at lower fan speeds may actually be so subtle as to be eliminated by some compensatory adjustment in creating a graph.
I myself use Silverstone NJ700, a fanless PSU, NH-P1, NF-A12x25 LS-PWM, and NF-A12x25 PWM now.
After posting the above comment, it occured to me that your ambient noise level might be determined by your air-conditioner. You might need to turn it on even if you don't normally have to in order to maintain a certain ambient temperature for testing. (At this time of the year, I do not turn on an air-conditioner thanks to the climate of the area I live in.)
@@osardegozar2581 nope. I turn that bad boy off and do multiple runs and compare results over runs to ensure I’m unrelated outside noise is not impacting results.
@@pc_analytics Then, given all the above, your PSU seems to be the only source of noise I can conceive of that determines your ambient level. EVGA SuperNOVA 850 GA may perhaps be close to 750 GA. Perhaps PSU fan makes a greater contribution to the ambient level in an open-air case.
Let me note that I do understand you've done everything you could with your current setup to ensure accurate measurement. I just want to make sense of noise levels of these coolers shown for a fan speed below 45%.
Thermalright Phantom Spirit 120 SE vs Cooler Master MASTERAIR MA824 STEALTH
Noctua ram clsarance is kne challenge. Yet low profile non rgb just fine
I prefer air cooler than Aio anh times
Simply blow external.air cooler in front of mesb panel who also has intake fan to blow air jnti the system.
By that thougb cosr a bit of power yet at the oing term I manage to remove the short curcuit incident. If the cooler air.from that air cooler is cold enougb and the wind is strong enough, even at maximum ooad no throttle
1:26 at last! 2 blinks in a row! thought i was watching at x1.75...
Oops, I meant to edit that out 😜
@@pc_analytics 😋
I would be curious on how they do at 250 watts
00000 proofs. where is picture ?
Check my community page. I think both coolers are in my pic of a wall of coolers
😎
A creepy man, looking into the camera without blinking, like a snake. Scary.
Why you so mean
it looks like he's got one of those a.i. programs that makes the eyes always point toward the camera
I intentionally cut out all my blinks just for you 😉
@@pc_analytics and the speed you flawless talk that down
MACHINE 🤣but exactly what i needed since i'm currently diggig through hundreds of reviews building a pc
are you a cyborg like Data from star trek? You dont blink at all.
Cyborg in training
Did you use the rev 4 offset brackets for the Arctic and the 7mm offset brackets for the Noctua?
I used the provided brackets for both of the coolers on the 5800x