Question Advice needed on Arctic Liquid Freezer III 360 - - - strange CPU temps ?

May 27, 2025
14
1
15
Hello everyone! I recently built a new computer. This is my first experience with a DIY build. Previously, I selected components and ordered assembly from a store.
Here’s the list of hardware:
CPU - Ryzen 9800X3D
MoBo - Asus X870-F Gaming
Case - Thermaltake CTE E660 with 6 140 mm fans (bottom and rear, intake) and 2 120 mm on top.
AIO - Arctic Liquid Freezer III 360, installed at the front side for exhaust to the right side.
GPU - MSI Ventus RTX 5080

I did not overclock either the CPU or RAM; in the BIOS, I only disabled Fast Boot and switched PCIE to Gen 4 because I am using a riser that only supports Gen 4.

I am very concerned about the CPU temperature, especially its dependence on tightening the pump screws.

I applied MX-6 paste, which came with the cooler, in a cross pattern, as indicated in the manual. I tightened the pump screws about 70%, I believe. A small amount of paste oozed out, so I assume there was enough paste.

After installing Windows and everything else, I launched Star Wars: Outlaws and got a CPU temperature of around 65 degrees Celsius. However, since I read the AIO installation instructions carelessly, I thought I had tightened the pump screws too much and decided to loosen them a bit. After that, I got around 53 degrees in the same game in the same scene, and I was using the all-in-one cable.

A day later, I decided to change the cable to CPU+Pump+VRM. I removed the pump from the CPU and saw that there was indeed enough paste. I wiped it off and reapplied it from the syringe, also in a cross pattern, but a bit less (as it turned out, that was also sufficient). Visually, I assessed that if I gathered the applied paste into a blob, it would form the recommended pea-sized blob. Then, still thinking that the pump screws should not be fully tightened, I tightened them about 30% with small turns alternately. The paste did not ooze out. After that, I configured the fans using Armory Crate. Here are my settings. C is degrees Celsius, and RPM is RPM.

CPU fan
1 - 20/20 C/RPM
2 - 30/25 C/RPM
3 - 40/34 C/RPM
4 - 50/45 C/RPM
5 - 60/60 C/RPM
6 - 71/76 C/RPM
7 - 80/100 C/RPM
8 - 100/100 C/RPM

Chassis fans (1, 2, 3 - case fans)
1 - 22/14 C/RPM
2 - 30/18 C/RPM
3 - 40/20 C/RPM
4 - 50/25 C/RPM
5 - 60/30 C/RPM
6 - 68/36 C/RPM
7 - 72/50 C/RPM
8 - 100/60 C/RPM

Chassis fan 5 (VRM fan)
1 - 20/20 C/RPM
2 - 30/25 C/RPM
3 - 40/34 C/RPM
4 - 50/45 C/RPM
5 - 60/60 C/RPM
6 - 67/80 C/RPM
7 - 70/100 C/RPM
8 - 100/100 C/RPM

AIO Pump
1 - 20/80 C/RPM
2 - 30/81 C/RPM
3 - 40/83 C/RPM
4 - 50/85 C/RPM
5 - 58/87 C/RPM
6 - 65/92 C/RPM
7 - 70/100 C/RPM
8 - 100/100 C/RPM

As you can see, if the CPU heats up to 80 degrees, both the pump and radiator fans run at 100%. Idle temps are 40 for the CPU package and 37 for the CPU, which is good. In Ghost of Tsushima, it's 49-50 degrees; in Assassin's Creed: Valhalla, around 55 in the forest and about 58 in a town; in Star Wars: Outlaws, from 52 to 57; in Cyberpunk, from 55 to 65 depending on the scene; in Oblivion Remastered, 50 indoors and from 58 to 68 outdoors. I believe these are good temperatures, but it should be noted that I used Frame Generation in games where it was possible to achieve stable 138 FPS (I have a 144 Hz monitor with G-Sync); without FG, the temperatures should be a bit higher.

It should also be added that when loading a game, for example, Ghost of Tsushima, the temperature could spike for a fraction of a second to 85 degrees, but as far as I know, this is normal and there's nothing that can be done about it. I should also mention that during the first launch of the game in RPCS3 while compiling, the temperatures were around 76 degrees.

In the Prime95 Small FFT test (I didn't change anything in the Run a Weaker Torture Test section), the temperatures were around 86 for the CPU package and 76 for the CPU. The test was conducted for 30 minutes, and around the 12-minute mark, the temperatures briefly spiked to 88 degrees. According to reviews on the Internet, I should have received lower temperatures even without undervolting. This kept bothering me, and I continued to search for information online. I accidentally discovered that the instructions explicitly state that the pump screws should be fully tightened. I saw the same information on some forums and in a video from Gamer Nexus. So yesterday, I fully tightened the pump screws. After that, I ran Prime95 again, and within 5 minutes, the temperatures exceeded 90 degrees; in Ghost of Tsushima, it's now 53 degrees, and at idle, it's 41 for the CPU package and 38 for the CPU. Yes, in games, the difference is small, but it exists! In Prime95, the situation is much worse. And it is about 85 when compiling stuff in RPCS3. Now I am completely confused and don’t know what to do.

In a day or two, a new MX-6 will arrive, and maybe I will try to repaste and remount the pump, and this time I will apply the paste in the center as one blob, although I doubt this will improve the situation, as when I removed it last time, it was clear that the paste had spread across the entire IHS.

I really need advice and information. I'm feeling very sad and confused. Any help is greatly appreciated!

PS Sorry for any mistakes if there are any. English is not my native language, but I am doing my best and using the help of AI.
 
Last edited:
Welcome to the forums, newcomer!

I really need advice and information. I'm feeling very sad and confused. Any help is greatly appreciated!

AIO - Arctic Liquid Freezer III 360, installed at the front side for exhaust to the right side.

Ideally that spot should be for intake, whether you decide to do push or pull on the fans is up to you but I'd flip the exhaust to intake and see if your temps change.

I only disabled Fast Boot
Make sure you're on the latest BIOS version and enable Fast Boot.

Reading through your thread, what's probably happening is that your constant tightening and loosening of the screws is causing an improper mount. Hot and cold spells will cause expansion and contraction for all things. You're advised to do everything in one consecutive process, which is preparing mounting hardware/surface, apply peas sized amount of thermal paste, applying the same mounting force on all corners in a criss-cross pattern.

A side note, use isopropyl alcohol to prepare both the CPU's IHS surface as well as the AIO's cooling block so you don't have residue from your prior thermal paste applications.
 
Thank you for your answer!


Yes, at the time of assembling the PC, which was on May 6 of this year, I had the latest version of the BIOS—I updated it myself right after the assembly and before installing Windows. I am using version 1303 from April 30.


I'm afraid that in my case, I won't be able to reinstall the AIO as exhaust because the fan cables are routed under the tubing, so I can't move them, and it's impossible to rotate the entire radiator due to the tubes. Just in case, I'm attaching a photo.

20250509-183141.jpg


Could you please explain why it's necessary to enable Fast Boot? I turned it off so that my computer would start completely fresh. I turn it off at night. Moreover, Fast Boot may affect that the Total reads and Total writes counters since the system starts are not resetting in HWInfo. Perhaps I'm mistaken.


In my original post, I described everything quite clumsily, so I will outline my tightening/loosening of the pump screws once again.

Initially, I tightened the pump screws (my AIO has only two) to about 70% because I didn't know they needed to be fully tightened.

A day later, I decided to loosen them a bit and got lower temperatures.

Later, I decided to change the AIO connection cable to separate ones for the CPU fans, pump, and VRM fan, so I completely removed the pump from the CPU and wiped the paste from the IHS and pump plate. After that, I applied new paste and tightened the pump screws to about 30%. After that, I got acceptable but uncomfortable temperatures of 86 degrees in Prime95.

After about a couple of weeks, I found out that the pump screws needed to be fully tightened, and I did that. After that, my temperatures worsened. Judging by your response, the deterioration could have been due to this, and the screws should have been fully tightened right when installing the pump.



When I receive the new paste and isopropyl alcohol, I will try to remove the pump, clean everything, and reinstall it, tightening the screws just once. BUT! I'm not sure that the pump screws should be fully tightened, even though it is clearly stated in the manual. The thing is, when I first installed the pump, I was afraid that I had tightened the screws too much (not knowing there isn't realy any "too much") and decided to check the back of the motherboard. I found that it was slightly bent at the CPU mounting point. There is a thick metal plate for reinforcement, and it is also slightly bent. This might be normal, but I am quite an anxious person and worry about new and expensive hardware. Can you tell me if this is how it should be? I definitely did not over-tighten the pump screws. With my AIO, it seems almost impossible to do so due to its design.


When I was wiping off the thermal paste, I used 90% medical alcohol; I didn't have anything else, and I read that this would work too. But now, along with the new MX-6, I will get isopropyl alcohol.


And one more thing. I'm concerned that the temperatures of the CPU and CPU package can differ by as much as 12 degrees at certain moments. I asked a neural network about this and received a response that it is acceptable :) But of course, one should be cautious with AI responses.
 
Are you using One for all cable from pump to CPU_FAN header on the MB.? It would be better to use the other one so you can monitor and control each part separately. With the single cable pump is regulated together with radiator and VRM fans and I noticed that it never reaches full speed. All headers on the MB should be set to PWM and to take signal from CPU temperature.
Where did you get that part about 70% for tightening screws ? Those on pump have a shoulder that limits tightening and should be tightened fully but not with excessive force, just to that shoulder.
PS. Looking at your picture, looks like all the fans are set to exhaust so the airflow is limited so you should turn those on the radiator to be intake as well as at least those 2 on the bottom under the GPU. As for BIOS, find settings for "Curve Optimizer", usually under PBO Overclock settings and set offset (all cores) to negative 20 for the beginning as most default settings in BIOS tend to overvolt ostensibly for stability's sake, with CO voltages are adjusted dynamically as need.
 
Last edited:
Hi, CountMike!

No, as I stated in the original post, rear and bottom fans are intake. They have reverse blades. And now i'm using cable with separate connectors for fans, pump and VRM fan. You can see curve settings in orignal post.

All fans headers monitoring CPU package temperature as it is higher than the CPU temperature. Except CHA_FAN5 - I plugged VRM fan there and set it to monitor VRM temperature.

About tightening pump screws. I didn't noticed that text in manual that says to fully tighten the pump screws, so I thought that I should decide by myself on how tight it must be. Now I have these screws tightenet fully but without excessive force.

Thank you for advice about Curve optimizer. I've tried to set this to -20. I ran Prime95 for a minute and didn't notice any changes in temperaturest. But I was nervous about all my issues, so I disabled CO. Now I want to resolve my issues with AIO and then I'm planning to play with CO and maybe RAM. You mentioned that automatically MoBo tends to overvolt. According to HWInfo, I have 1.225 V CPU core, 1.021 V CPU SoC. And my PPT limit is 90%, so during Prime95 test CPU draws about 145 W oe power. I'm not sure, but it look okay. And these are maximum values, not constant.
 
Last edited:
What is your primary goal here.

A very important thing to know is the actual temperature does not matter as long as the CPU does not reduce it clocks because of the heat. For your CPU as long as it does not get to 95C the clocks run at the maximum rate. So do not stress over if you get 85 or 87 or even 90. Hwinfo fairly clearly shows if you are thermal throttling.

It is more a trade off of the noise the fans make and the temperature. As long as you do not thermal throttle you are good.

The main issue you tend to see is GPU is what makes the most noise.
 
Hi, CountMike!

No, as I stated in the original post, rear and bottom fans are intake. They have reverse blades. And now i'm using cable with separate connectors for fans, pump and VRM fan. You can see curve settings in orignal post.

All fans headers monitoring CPU package temperature as it is higher than the CPU temperature. Except CHA_FAN5 - I plugged VRM fan there and set it to monitor VRM temperature.

About tightening pump screws. I didn't noticed that text in manual that says to fully tighten the pump screws, so I thought that I should decide by myself on how tight it must be. Now I have these screws tightenet fully but without excessive force.

Thank you for advice about Curve optimizer. I've tried to set this to -20. I ran Prime95 for a minute and didn't notice any changes in temperaturest. But I was nervous about all my issues, so I disabled CO. Now I want to resolve my issues with AIO and then I'm planning to play with CO and maybe RAM. You mentioned that automatically MoBo tends to overvolt. According to HWInfo, I have 1.225 V CPU core, 1.021 V CPU SoC. And my PPT limit is 90%, so during Prime95 test CPU draws about 145 W oe power. I'm not sure, but it look okay. And these are maximum values, not constant.
So, it looks like you did everything right so far. That's strange to have such temperatures, high although not excessive and within CPU limit. When I run into a problem like that first I set all fans, pump etc at full speed in an open case to eliminate air and liquid flow problems. That usually leaves just cooler to CPU contact as a problem. It's possible that CPU or cooler are not flat. If I apply just small and thin amount of paste, I can see it's footprint indicating some uneven surface.
PS: My LF III 360 keeps my well performance optimized 7900x up to 80-82c (Tjmax is 95c) under full load for hours at which it draws 170w.
 
Hey, bill001g! Thank you!

I cannot fully agree with you. I'm reaching 90 degrees Celsius with PPT limit 90 %, wich is 145 W in my case. If my CPU will draw full 162 W of power then my temperatures will rise significally and there will be thermal throttling. But yes, now in real tasks (gaming and Stable Diffusion) CPU temps are well within thermal limits.
And I don't have noise issues. My AIO fans are okay even at 100 % speed. Same for GPU. The pump and maybe VRM fan are a bit unpleasant, but it is not really a problem.

My problem is that my AIO should perform much better according to its characteristics and user's report. As an example, CountMike has the same AIO and with 170 W reaching just 82 degrees. That is brilliant performance! And another problem is that I have slight bent of MoBo in the spot where CPU socket is located.


CountMike, I've tried to remove glass panels and I also tried to leave panels but run all case fans at 100% (it was LOUD). An I didn't noticed significant changes in temps, but it showed that airflow in my case is good. So, I believe, this is not a problem.
Can you tell me something about MoBo bending? We are have same AIO. Do your MoBo slightly bent? Maybe it is normal, I don't know. It scares me, because I installed pump right in terms of tightening the screws.
When I removed pump to change cable from all-in-one to 3-headers I saw that paste was on all IHS's surface. I'm not experienced, but for me the print looked adequate.
What is best pattern to apply thermal paste? I applied in cross-shape because in manual this was pointed as best approach. Maybe I should try to make a blob in the center?
 
Last edited:
Hey, bill001g! Thank you!

I cannot fully agree with you. I'm reaching 90 degrees Celsius with PPT limit 90 %, wich is 145 W in my case. If my CPU will draw full 162 W of power then my temperatures will rise significally and there will be thermal throttling. But yes, now in real tasks (gaming and Stable Diffusion) CPU temps are well within thermal limits.
And I don't have noise issues. My AIO fans are okay even at 100 % speed. Same for GPU. The pump and maybe VRM fan are a bit unpleasant, but it is not really a problem.

My problem is that my AIO should perform much better according to its characteristics and user's report. As an example, CountMike has the same AIO and with 170 W reaching just 82 degrees. That is brilliant performance! And another problem is that I have slight bent of MoBo in the spot where CPU socket is located.


CountMike, I've tried to remove glass panels and I also tried to leave panels but run all case fans at 100% (it was LOUD). An I didn't noticed significant changes in temps, but it showed that airflow in my case is good. So, I believe, this is not a problem.
Can you tell me something about MoBo bending? We are have same AIO. Do your MoBo slightly bent? Maybe it is normal, I don't know. It scares me, because I installed pump right in terms of tightening the screws.
When I removed pump to change cable from all-in-one to 3-headers I saw that paste was on all IHS's surface. I'm not experienced, but for me the print looked adequate.
What is best pattern to apply thermal paste? I applied in cross-shape because in manual this was pointed as best approach. Maybe I should try to make a blob in the center?
There should be no fear of MB bending, back plate is sturdy enough to prevent it. The CPU socket and clamps hold it on all sides unlike Intel so that's safe too. That's why Arctic provided additional clamping (contact frame) for Intel and not for AMD. I did install additional Contact frame instead of original sheet metal but I don't think it does much.
As for applying paste, I always preferred careful spreading specially thick and hard ones like Mx-6. I cover as thin as possible but enough not to see heat spreader writing, also leave about mm or two around the edges of the CPU, that's enough to prevent spilling out. Taking time to do it carefully pays back later on. There is another system some use, "Screen print", that consist of a thick mesh like that one for windows bug screen, put a piece over CPU and apply paste over it, that leaves enough paste in checkered pattern and under pressure it all fills up nicely without any air baubles.
 
But in my case MoBo and built-in metal backplate are slightly bended. Is it normal or it must be cimpletely flat? I'm trying to understand, is it something wrong with my MoBo or not. Backplate is thick and made of metal. One the one side I doubt I could bent it when tightened pump screws, on the other there is a fact that it not flat. I don't understand what is going on 🙁

I've checked my previous PC with ASUS Strix Z590-F MoBo and Fractal Design Prisma AIO. It also has some bend. Maybe it is really normal thing?

Thank you for info about paste application. Maybe I'll try to spread it or just apply a blob on the center of IHS. If I can't apply ideal amount of paste, then what is better: use less or more? Of course, within reasonable range :)
 
Last edited:
But in my case MoBo and built-in metal backplate are slightly bended. Is it normal or it must be cimpletely flat? I'm trying to understand, is it something wrong with my MoBo or not. Backplate is thick and made of metal. One the one side I doubt I could bent it when tightened pump screws, on the other there is a fact that it not flat. I don't understand what is going on 🙁

Thank you for info about paste application. Maybe I'll try to spread it or just apply a blob on the center of IHS. If I can't apply ideal amount of paste, then what is better: use less or more? Of course, within reasonable range :)
Motherboard should be perfectly flat and back plate is there to ensure it's flat and can withstand pressure at least around CPU socket. In any case that little bend it may have, can't cause cooler to have no adequate pressure. If MB is too bent, it would brake some traces (there are several layers of them) so heat would be least of a problem.
At this point, I would be inclined to thing something may be wrong with the cooler, There was a batch of LF II that had faulty gasket on the pump which restricted liquid flow and Arctic offered replacement or a repair kit. Could be something in that effect.
Since you have MB rotated by 90 degrees pump is sitting sideways although pointing only way it can to MB. You might want to contact arctic to see what they think or maybe ask for replacement, possibly for new Pro version. It's supposed to be better by 5-10c on Razen but little on Intel.
Have a look at this video.
View: https://www.youtube.com/watch?v=oi8Ic8DtLQg
 
Yes, it is logical that MoBo has to be flat. But my isn't 🙁

My AIO is revision 3, if that matters. I'll watch video later and try to contact Arctic. Also I should ask them about bending. Thank you for trying to help!

Ah, yes, I remembered. During the first few power-ups of the PC right after assembly, I heard the sound of liquid sloshing around for a while. But those sounds quickly disappeared and haven't been heard for a long time now. Could this be a sign of something bad or good? In the video you suggested it is stated that with "tubes on top" mount some buble noise can be heard. Maybe it is what I heard?
 
Last edited:
In closed loop systems there has to be some air because of liquid expansion when it heats up, otherwise something could explode or spring a leak from pressure. When it's sitting in a box, air would go to top position so yeah, there could be some gurgling sound until it evens up although I don't remember hearing it. Once it gets to working temperatures, air gets compressed so there's little of it. Many say that hoses should enter radiator at bottom so pump doesn't suck in any air while cold and that it cools a bit better. I found very little difference, 1-2c tops.
Prior to LF III, I had 360 LF II on Ryzen 5800x (also hot CPU) and 240 first LF on Ryzen 3700x. Prior to those I had Cooler Master Nepton 140 XL with push-pull fans on my AMD FX 8350 which tended to run hot to. It was more than adequate allowing 4.9GHz OC..
 
Wheew!.. I received a fresh MX-6 and 99% isopropyl alcohol. I checked the paste using the QR code - it’s genuine, which is not surprising since I bought it from a reliable store.


So, I removed the pump and carefully wiped the paste off the IHS and pump plate. Unfortunately, I didn’t take a photo of the imprint, but it looked normal, though I have little experience.


Then I applied the new paste. I decided to apply it in a cross shape again. I also decided to add a little in the CCD area. After that, I carefully installed the pump, tightened the pump screws a bit, and then tightened them a quarter turn in sequence until I felt a noticeable increase in force. I was nervous, so I didn’t check for MoBo bending, and now I don’t want to remove the case wall again, sorry.

20250528-205145-1.jpg


Now, after some light testing, I have 39 degrees Celsius for the CPU package and 35 for the CPU. Here are the test results. What can I say... They are better than when I just re-tightened the pump screws. In my opinion, the result is not the best, of course, but it can’t be called bad either. Especially considering that I didn’t adjust anything in the BIOS, meaning all voltages and limits are in auto mode and there’s no undervolting. And the RAM, by the way, is also not overclocked.

Fully-tightened-screws.png


Please share your opinion on the temperatures.


Thank you for the information about the reasons why there is air in the AIO. So, everything is fine on that front.
 
RPCS3 is a Playstation 3 emulator. When you run a game for the first time (or if you cleared compiled stuff before launch) then it has to compile some thing. So, yes, it is real task with high CPU load.
Just finished Cinebench multicore. What do you think about results? I really don't like that 91.8 spike during processing textures...

It is worth mention that during test (and Prime95 as well) the radiator was just slightly warm. And it is noticable more warm while gaming for an hour or longer with constant 55 degrees Celsius. Is it normal?

Cinebench-temps.png
 
I am not sure what your issue is. Are you just concerned that your numbers are higher than other people you see on the internet.

The flaw with comparing temperatures numbers directly between different testing locations is the room temperature you start with can be very different. If pc had a room temperature sensor and Hwinfo would auto correct these numbers you could directly compare them.

Sites that do cooler comparisons are very careful to make sure the temperatures of the room they are testing in are exactly the same. Even small difference can magnify difference between coolers ruing the results.

I don't see your problem. It should not affect your usage. The thermal throttle line show "NO" all the way across.
 
From my perspective, high 80's with ~145 W power draw is very high, taking in account that I have many case fans and high performance AIO. And wattage is only 90 % of what 9800X3D can draw. If I remove power limit then I'm pretty sure my CPU will reach 95C at 162 W.
The ALF III 360 can dissipate from 200 W to 300 W if this info is true. So my "humble" 145 W should be a light work for AIO. Look at CountMike's post #7 - his CPU draws 170 W and has only 82C at full load and he has sasme AIO.
Yes, room conditions matters. I've tried testing my PC with about 25C ambient temperature and with AC turned on. With AC my room was, I dont' know, about 18-20C, it was uncomfortable. So, I think I should had better results, but no.
On the other hand, I've read that 9800X3D is pretty hot CPU even if it draws relatively small power at the moment. Maybe it is due to 5 nm.
It is also possible that I get these temperatures because my pump's tubes is horizontal because MoBo is rotated 90 clockwise. with "classic" hardware mount tubes wuld be facing down. Maybe I should email Arctic and ask about that.

If I'm wrong, please, share some information.

Oh, and just for interest. My idle power draw is about 17 W. How hot would be 9800X3D if there is no colling at all? I've asked ChatGPT (or Qwen, I don't remeber) and it said that CPU would overheat very fast.
 
You pretty much at the point you replace the AIO if you feel it isn't actually cooling like it should.

AIO are almost impossible to troubleshoot and no parts can be replaced.

Maybe pick up a cheap but effective air cooler like thermaltake peerless assign. You can get these for $35 on amazon most the time. These have been tested and can cool 200 watts.

I suspect you went with a AIO for appearance reasons so you would only use this cooler to test. It would quickly tell you if it really is the AIO and you would have something to use if you needed to RMA the AIO.
 
Sorry, but I don't understand you. First you are saying that there is no problem, but now you are talking about replacing AIO and testing temps with air cooler...

No, I went with AIO because of less compatibility problems overall and better performance (that was my thoughts). And I saw controversy opinions about ability of air coolers to cool such powerful CPU's.

Anyway, your advice is interesting. But there are several models of Peerless Assassin. If I decide to try it, wich model should I get? And please, look at photo of my PC. Is it possible to install an air cooler? I think it would be difficult because of GPU.
Damn, my old PC in 2013 was 3770K overclocked to 4.2 GHz and giant Noctua NH-D14. And it was ideal - zero issues!
 
It all depends on what you call a "problem". If the cooler keeps the cpu under the throttle temperature then it is doing its job. Now if you think the number should be lower because other people get lower numbers with a similar setup then that is more of a problem where you feel you did not get what you paid for.

This is a older review but you will see similar results from pretty much everyone.

https://gamersnexus.net/coolers/cha...erless-assassin-cpu-cooler-review-benchmarksh

There are newer version of this cooler (same name but they have V1,V2 adn v3) it seems they are minor fan changes and increased RPM. You trade noise level for slightly more cooling.

Like any air cooler or even AIO you need to carefully measure things to see if it will fit. The height tends to be the largest issue for most air coolers.

This peerless assassin model overlap the memory a bit and will go over the top of the VRM shield. In your photo I would be a bit concerned if the power cable for the video card would touch the fan. Depends how much you can move the cable around. Depends if you are going for long term or just test install.

If you did not already have a AIO I would have just said go with the air cooler. In this case I would use the peerless assign mostly to prove to the AIO vendor that the AIO is defective in some way. If you look at the charts in the review all 360 aio when you run the fans at 100% beat the peerless by aleast 10%. Yours should also.

Just remember all these coolers have no issue cooling a 200 watt cpu espeically if you are willing to run the fans at 100%
 
Think about this more your motherboard is strange with the memory on the side. I am not sure what direction the peerless would sit. Normally it is designed so you can move the fan up a tiny bit if it does not clear the memory. If you were to run the cooler with fans front to back then the side of the cooler would be above the memory. Not sure how much clearance there is for that. Running the cooler with the fans up and down it would likely sit above the memory but I am unsure how that relates to how the CPU chip is oriented.
 
My motherboard is rotated 90 degrees clockwise from "classic" orientation. So will be with cooler. Its outer fan will be above RAM. The main issue will be GPU cable. If i decide to buy PA then I'll have to bend the cable pretty much or buy a third party cable with L-shaped connector. Outer fan of the cooler will most likely need to be raised above RAM, but I don't think this will be an issue. As for cooler's height, my case supports up to 166 mm and, for example PA 120 is 157 mm.

Again, can you please tell wich Peerless Assassin I should buy if I decide to. There is 90, 120, 140 versions, and also an "SE" version?

And by the way. I saw a discussion on another forum where a user has similar problems with same CPU and AIO. And he got 86-87C in Cinebench and he has -30 CO with 200 MHz boost and 10x Scalar. So either we both have bad AIO or this is a normal performance (but I doubt).
 
Last edited:
Thank you for information!

Meanwhile, I contacted Arctic support, but have no answer yet.

I've also tried to enabled PBO Enhancement Level 2 (80C). In Cinebench my scores are same or sometimes 5-10 point higher than stock. My temp are 70C at most. And in Prime95 Small FFT test my temps are 80C and clock speeds increased by 100-200 MHz.
In Cinebench power draw was 125W, now 90-95W, in Prime was 145W with lowering to 130W, after tweak it is 130-135W.
Does that mean its just pure benefit - same performance and much lower temps and power draw? These results are so good to be true...