AIR vs AIO. 18c drop in temps

[Robert896r1]

Making a small list of people who don't know how to setup cooling configurations according to The Last Airbender.
- Derbauer
- Buildzoid
- Kitguru
- TechYEScity
- Craftcomputing
- Optimum tech

Don't worry, I'll try to keep the list updated for you.

 

[doyll]

Making a small list of people who don't know how to setup cooling configurations according to The Last Airbender.
- Derbauer
- Buildzoid
- Kitguru
- TechYEScity
- Craftcomputing
- Optimum tech

Don't worry, I'll try to keep the list updated for you.

Could you post links to hwo each do their testing? Then we can discuss each on it's own merits.

Edit: just looked and only one that has a physiclal website doing text reviews is Kitguru. All the rest are youtube taking heads or facebook media, so you are probably right, they don't know how or at least don't run cooling tests properly.

It's time you stop trolling and act like a contributing member of forum.

I posted data supporting what I'm saying but you post your own opinion and names, nothing supporting what you say.

 

[Robert896r1]

I've posted a large sum of data and reference points from my testing and that of others who are well respected. You're drunk off confirmation bias. I, nor anyone else can change that. Report away or create your own thread from legacy data points that have little relevance.

 

[doyll]

KitGuru:

Testing Methodology: Thermals/Acoustics
To measure the performance of CPU coolers, we devised an easily repeatable series of tests. The only variable is the coolers themselves to ensure the results can be accurately compared against one another.
Test Rig
To reflect current hardware demands, we have updated the test system at KitGuru. It now comprises of a Z390 platform and included is the Intel Core i9 9900k CPU, mated with a Gigabyte Z390 Aorus Master motherboard and 16GB (2x 8GB modules) Patriot Viper RGB running at 3600mhz CL16. Our storage device is a Corsair Force LE 120GB SATA SSD. Power to the test system is managed by a Seasonic Platinum Series 1000w PSU.


Thermal Testing Procedure
The procedure will consist of several tests, that will produce six temperature readings for each cooler. The data can then be used to compare cooling performance.

First, we will lock the 9900k at its 3.6GHz all-core base clock speed, with a 1.15V Vcore applied in the BIOS. Then, let the system idle in windows for 15 minutes and measure temperature. At the same frequency, we will then measure the load temperature.

To simulate overclocked frequency, we will then lock the 9900k to 4.7GHz on all cores with a 1.20V Vcore and run the same idle and load test again. To measure an extreme overclocking situation, we will lock the 9900k frequency to 4.9GHz on all cores with 1.30V Vcore and run the tests one final time – this should push the limits of the cooling performance of even the very best CPU coolers.

Admittedly, the overclocked frequencies/voltage may not be the highest achievable, nevertheless, it will give us a good set of data to compare the effect frequency/voltage changes have on temperature.

The temperatures presented in the graphs will be average Delta temperatures measured over the length of the test. We will calculate the Delta temperature by deducting the ambient temperature in the test room from the measured CPU temperature, both at idle and under load.

All idle temperatures are measured by letting the system idle in Windows for 15 minutes. Load temperatures are achieved by running AIDA64 stressing CPU, FPU and cache again for 15 minutes. This should give the CPU enough time to reach its maximum temperature. Throughout testing, Load Line Calibration will be set to level 6 (turbo) in BIOS and all CPU Fans/pumps set to maximum RPM.


Noise Testing
Noise levels will be measured with our sound meter at a distance of 1 meter from the test bench. The cooler’s fans will be set to their maximum RPM rating. Ambient noise level in the room is measured at 37.5dBA and the peak noise level of the cooler will be recorded in dBA.

Nothing in their testing methology saying anything about when, where & how temps are taken, if airflow temp entering cooler/radiator is taken at same time as CPU temp, or if noise levels are equalized during tests.

 

[doyll]

I've posted a large sum of data and reference points from my testing and that of others who are well respected. You're drunk off confirmation bias. I, nor anyone else can change that. Report away or create your own thread from legacy data points that have little relevance.

But none comparing your setup as it is now against same system with top tier air cooling .. all making your lame claim of CLC being 18c better your Dark Rock 4 with performance similar to Hyper 212 at similar fan speed.

 

[Ross Thomson]

Only in your mind, not in reality.
I'll make you the same offer; bring your sytem over and I'll show you how it will perform as well on top tier air as it does with your CLC.

If by my mind you mean the internet then yes. I've not seen one review where an air cooler tops the table of cooler reviews.

I don't have a CLC and an air cooler doesn't do as well as an AIO/CLC on a ryzen 3900x anyway.

 

[Ross Thomson]

Making a small list of people who don't know how to setup cooling configurations according to The Last Airbender.
- Derbauer
- Buildzoid
- Kitguru
- TechYEScity
- Craftcomputing
- Optimum tech

Don't worry, I'll try to keep the list updated for you.

it's basically anyone that he doesn't agree with is wrong, doesn't do things right or the results are not valid because it was done in a case. I've blocked him on other sites because of his nonsense and done it here now.

 

[doyll]

iIf by my mind you mean the internet then yes. I've not seen one review where an air cooler tops the table of cooler reviews.

I don't have a CLC and an air cooler doesn't do as well as an AIO/CLC on a ryzen 3900x anyway.

No, it's in your mind and you haven't posted any credible links to data supporting your claims.

AIOs are slightly better than top tier air, CLCs have a very few that are as good or few degrees better than top tier air. I've posted data and links backing up that claim .. but your posts are all talk and no supporting data and link.s

it's basically anyone that he doesn't agree with is wrong, doesn't do things right or the results are not valid because it was done in a case. I've blocked him on other sites because of his nonsense and done it here now.

Not at all! It's only those that post up statements that they can't back up with anything but what they claim.

So you post up more false insinuations about me instead of verifiable data showing what you claim. How magnanimous of you.

So oh great one, it's time you either give us proof of claims.

I've posted how to do accurate testing, how about you tell us what you think is an accurate way to test air coolers, AIOs, and CLCs.
Tells us if airlfow temp into each and equal noise level should or should not be part of criteria used in comparison testing.

 

[Cyber-Mav]

Most reviewers Dont use a case and have open test bed

 

[jeff83nz]

I'm don't know much about a Gaming 5 so can't comment what the right llc is but on the Aorus series, the vdroop on high is just too much and makes you run a decently high idle to make up for it. On my board, Turbo is only level 3 of max. Extreme and Ultra Extreme are above it. I don't use those.

I had a Aorus Gaming Ultra. Spec wise for the money read great, but has completely put me off them. Switching to what I am on now (Asrock Phantom Gaming-ITX) has given me a much more stable O/C, at a lower vcore and thus much lower temps. It meant I could go from 4.7 all cores to 5.2 (avx offset currently 2).

As for this unending war of air vs. aio inc clc.

Have a CLC 280mm and a Noctua D15. The main advantage in my view of CLC is it is easier to fit the graphics card as the D15 is generally massive, and they look nicer. D15 though is quieter when not under load due to lack of pump noise, and quieter overall if you do not upgrade the fans on the CLC as most are pieces of noisy ****. With adequate case ventilation both will do the job 99% will need. AIO is likely to more ensure you have good case flow though.

One exemption is SFF. In this case AIO / CLC is definitely the way to go, although the TU150 is the exception

 

[doyll]

Most reviewers Dont use a case and have open test bed

More and more review sites have changed from cases to bench test stations, but I wouldn't say 'most'. While testing on open bench does help, there can (and is) still many variables not delt with like monitoring cooler/radiator intake air temp and recording it at same time as CPU temp is. The air temp into an air cooler mounted on motherboard is not the same as it is into a radiator setting off to the side of motherboard. Many reviewers also don't run specific DB level to CPU temp tests either, and that is possibly the most important on to most of us because it tells us how well component cools at a given noise level.

I had a Aorus Gaming Ultra. Spec wise for the money read great, but has completely put me off them. Switching to what I am on now (Asrock Phantom Gaming-ITX) has given me a much more stable O/C, at a lower vcore and thus much lower temps. It meant I could go from 4.7 all cores to 5.2 (avx offset currently 2).

As for this unending war of air vs. aio inc clc.

Have a CLC 280mm and a Noctua D15. The main advantage in my view of CLC is it is easier to fit the graphics card as the D15 is generally massive, and they look nicer. D15 though is quieter when not under load due to lack of pump noise, and quieter overall if you do not upgrade the fans on the CLC as most are pieces of noisy ****. With adequate case ventilation both will do the job 99% will need. AIO is likely to more ensure you have good case flow though.

One exemption is SFF. In this case AIO / CLC is definitely the way to go, although the TU150 is the exception

NH-D15 is very wide and with no offset it's fin pack reaches 75mm toward PCIe so can be problematic. I don't know why Noctua took so long before marketing NH-D15S with it's offset base. Thermalright and others have been making offset base coolers for long time. Air being quieter than CLC is one of the biggest reasons I don't like them .. that and higher cost, can't repair, don't last as long, etc. Good case airflow is a given, doesn't matter if it's air or water cooling. Indeed, AIO/CLC installation often adds their fans to what case had before and of course this increases case airflow.

 

[Robert896r1]

To show the difference between my Alphacool 360LT AIO and what a custom loop allows:

This is me running P95 NON AVX for 90mins. CURRENT IOUT is what you want to pay attention to. The limit of my cooler is around 150amps average power draw as can been seen here:

Now Let's look at a custom loop 9900k running P95 AVX. You can see the custom loop allows for another 50amps (CURRENT IOUT) worth of cooling on average under sustained high loads.




The 9900k is a pain to cool remains true

 

[andy_mk3]

A good example that heat output does not equal temperature though. It's about how much heat you can remove.

I'm still blown away by my custom loop.

 

[Robert896r1]

A good example that heat output does not equal temperature though. It's about how much heat you can remove.

I'm still blown away by my custom loop.

? It's relative to demonstrating the thermal density and the cooling challenges of a 9900k.

 

[Stu]

KitGuru:

Nothing in their testing methology saying anything about when, where & how temps are taken, if airflow temp entering cooler/radiator is taken at same time as CPU temp, or if noise levels are equalized during tests.

The review states:

"The temperatures presented in the graphs will be average Delta temperatures measured over the length of the test. We will calculate the Delta temperature by deducting the ambient temperature in the test room from the measured CPU temperature, both at idle and under load.
Noise levels will be measured with our sound meter at a distance of 1 meter from the test bench. The cooler’s fans will be set to their maximum RPM rating. Ambient noise level in the room is measured at 37.5dBA and the peak noise level of the cooler will be recorded in dBA."

Why is this method not good enough for giving a fit for purpose comparison of two coolers?

 

[doyll]

The review states:

"The temperatures presented in the graphs will be average Delta temperatures measured over the length of the test. We will calculate the Delta temperature by deducting the ambient temperature in the test room from the measured CPU temperature, both at idle and under load.
Noise levels will be measured with our sound meter at a distance of 1 meter from the test bench. The cooler’s fans will be set to their maximum RPM rating. Ambient noise level in the room is measured at 37.5dBA and the peak noise level of the cooler will be recorded in dBA."

Why is this method not good enough for giving a fit for purpose comparison of two coolers?

I missed that part, but they are figuring delta from test room temp, not actual air temp into cooling component. The temp into cooler/rad on a test bench is likely not the same as elsewhere in room, and air cooler on motherboard will have differnt intake air temp then rad hanging off to the side. So testing ends up with something like +/-2c (total 4c) possible varience in air temp into cooling component, maybe even more.

 

[Stu]

I missed that part, but they are figuring delta from test room temp, not actual air temp into cooling component. The temp into cooler/rad on a test bench is likely not the same as elsewhere in room, and air cooler on motherboard will have differnt intake air temp then rad hanging off to the side. So testing ends up with something like +/-2c (total 4c) possible varience in air temp into cooling component, maybe even more.

Your description above indicates to me that you are interested in the absolute cooling performance of a cooler in isolation if you are trying to remove differences of air temp around a test bench versus air temp above the motherboard... while your approach makes for a good academic investigation, I don't think it is a good way to consider performance in a real system... a real system does have different temp air going into a radiator setup as an intake compared to an air cooler inside a case. I think you are too focused on minute detail and variability rather than real world application or a method that is fit for purpose.

 

[doyll]

Your description above indicates to me that you are interested in the absolute cooling performance of a cooler in isolation if you are trying to remove differences of air temp around a test bench versus air temp above the motherboard... while your approach makes for a good academic investigation, I don't think it is a good way to consider performance in a real system... a real system does have different temp air going into a radiator setup as an intake compared to an air cooler inside a case. I think you are too focused on minute detail and variability rather than real world application or a method that is fit for purpose.

You are correct, I want to know just how good the cooler (be it air or water) is. I don't think it's 'academic' to want as few variables as possilbe in cooler test results. Even using actual air temp into cooler unless testing is done on a synthetic heat sourcewe still have differnce is IHS to cooler seat between CPUs, and difference in CPUs even if same model. And it's not hard to eliminate air temp variables, so why not? At least my way we eliminate several degrees which can be a significant variable.

After all, what is your 'real world'system? Is it your, mine or someone elses? Every system is at least a little differnt from others. If test results are based on air temp into rad/cooler we at least are eliminating what we can to determine how well they perform against each other. No variable except CPU TDP and seat heat transfer. This way if we are getting 5c or 10c hotter temps than review did on same CPU / same TDP heat source it is fair to assume something isn't right and we can then re-seat cooler, adjust our case airflow to improve / lower air temp into cooler etc. Just so you know, I routinely build systems that air temp into cooler is 3c or less above room ambinet .. and can get it down to less than 1c if I take a little more time adjusting fan curves.

 

[Robert896r1]

Thanks purely to cold weather and offset voltage:



Basically if you can drop your ambient and start at a lower temp, that benefit extends to lowering your peaks thus giving you more headroom. In this case, my starting temp was 19c thus I was afforded the extra headroom. This is a "winter" config. Actual gaming and general usage temps are obviously a lot lower than this quick P95 run. Referencing earlier results, a drop if 5-7c nets me another 12amps average of heat output which is what I'd expect.

 

[unkzilla]

Hi all,
Bit of a random bump but this thread helped me with my decision to buy a H150i Pro and upgrade from a Noctua NH-D15 - thought I'd post my results

I'd consider my setup one of the best cases for air prior to the upgrade - 3x 140mm intake, 1x 140mm rear exhaust, 2x 140mm top exhaust, Noctua NH-D15 + 2x 140mm CPU fans (8x 140mm in total). Case is a master case pro 5 with mesh front and top panel and absolutely no clutter in machine - no drive cages or anything like that

9900ks +NH-D15 at 1.4v (LLC5/XI Gene) - 96c max, 93c average
9900ks +Corsair H150i PRO - 86c max, 83c average

Quite happy with 10deg drop and will re-purpose the Noctua for one of the Ryzen machines we have

https://postimg.cc/TpbRWr56