Review Arctic Liquid Freezer III AIO Review: To put it bluntly, the Liquid Freezer III is unimpressive

[Albert.Thomas]

Yeah, criticism. But calling the results worthless, ALL test results worthless, and questioning whether you fundamentally understand what you are doing goes beyond criticism. Accusations of publishing misinformation as well.

That's where it went too far. Questioning results is fair, esp when coolers underperform or strongly outperform rivals (Peerless Assasin, Phantom Spirit EVO).

I guess I must have a thicker skin. His comments were mild compared to those on Reddit

 

[Heiro78]

I had the 420 White ARGB and it was working really well with some lian li AL140 V2 in a push/ pull intake position at the front of my Phanteks G500A case. Unfortunately after 2 weeks of debugging in my free time I realized that ARGB in the VRM fan head wasn't working because of a stuck pin on the pump block. I returned it but now the 420s are sold out at their ebay store. I don't wanna go with the 360, and this article gives me further reason not to

 

[RubberMallet]

I've read the authors response on reddit and it doesn't make sense. I think the tepid response from some other outlets gave him an opportunity to push his agenda using this product as the scapegoat.

I would imagine that 90% of TH readers are on air coolers and feeding their confirmation bias is going to do more good for TH readership that publishing the facts.

A change in editor, and/or lack of screening contributors could drastically change the credibility of an outlet over night. We ve seen it happen to other outlets.

There is absolutely no way that a 360 aio set up properly on a properly functioning 13700K will thermal throttle at 235Watts. Other testers with 13900/14900K show it cooling 300W.

Hopefully this is just an honest mistake and he will figure out what he did wrong.

 

[thestryker]

There is absolutely no way that a 360 aio set up properly on a properly functioning 13700K will thermal throttle at 235Watts. Other testers with 13900/14900K show it cooling 300W.

Hopefully this is just an honest mistake and he will figure out what he did wrong.

I know responding to you is a waste of time as you're just trolling at this point. You're so stuck on wattage you either cannot or refuse to see reality.

I said it before but Albert uses an actual case in testing rather than an open bench and silicon varies. If you look at the results on this chart and compare them to say TPU you'll find his 13700K is around 60W lower on most results. This does not change the comparative accuracy of the results. TPU has only reviewed the 240 LFIII, but what testing overlap there is shows the same coolers in roughly the same place despite the wattage disparity. This is the only relevant part when testing a cooler because no reviewer's platform can ever mirror one in the wild unless you literally took those components from them and tested in an identical environment.

 

[Heiro78]

I've read the authors response on reddit and it doesn't make sense. I think the tepid response from some other outlets gave him an opportunity to push his agenda using this product as the scapegoat.

I would imagine that 90% of TH readers are on air coolers and feeding their confirmation bias is going to do more good for TH readership that publishing the facts.

A change in editor, and/or lack of screening contributors could drastically change the credibility of an outlet over night. We ve seen it happen to other outlets.

There is absolutely no way that a 360 aio set up properly on a properly functioning 13700K will thermal throttle at 235Watts. Other testers with 13900/14900K show it cooling 300W.

Hopefully this is just an honest mistake and he will figure out what he did wrong.

You know it's ok to have outlier information? Not every review has to conform to a concensus.

 

[Albert.Thomas]

I would imagine that 90% of TH readers are on air coolers and feeding their confirmation bias is going to do more good for TH readership that publishing the facts.

I'm afraid that the only confirmation bias going on here is from folks who are struggling to accept that a mediocre liquid cooler only performs on par with a good air cooler. There are strong liquid coolers on the market, but this ain't one of them.

If you doubt my results, feel free to recreate the conditions:

BeQuiet Silent Base 802 case, system fans set to low, mesh front panel and top used.

2x8gb G-Skill TridentZ DDR4-3466

MSI Z690 A-PRO Motherboard. "Out of the box" settings. Water cooler preset used when initially loading BIOS. Other than that, XMP is enabled for DDR4.

Intel i7-13700K CPU

23C Ambient Temperature

 

[Papusan]

Unfortunately, the Liquid Freezer III fails our tests and melts under the heat of Intel’s i7-13700K.

Arctic Liquid Freezer III AIO Review: To put it bluntly, the Liquid Freezer III is unimpressive : Read more

Hmmm. Thanks for the review.

What I don't understand.... Many reviewers have also tested this new AIO with new modern Intel platform but with different results. Who's right... Your or their reviews? Can it be a faulty products you got for the review? And then you have contact pressure with this new contact frame. Done pressure tests? Thanks

 

[risenres]

I dont get it. I have it and works great (on I7 12gen it gives average 10C less than Dark rock 4 pro). Every test (e.g. from Gamers Nexus) shows that version III is better than II (which you gave 4.5 stars). You didn't remove the protective sticker from the CPU heatsink, or something?

 

[helper800]

You didn't remove the protective sticker from the CPU heatsink, or something?

That would be funny. The first time I installed a LF II onto a friends build I forgot to peel that plastic myself. The CPU would slowly creep to 100C on a 12700k about 5 minutes after turning the PC on. Luckily I figured it out after about 10 minutes.

 

[Albert.Thomas]

What I don't understand.... Many reviewers have also tested this new AIO with new modern Intel platform but with different results. Who's right... Your or their reviews?

There are certainly a few reviews out there which I feel have subpar testing methodology. For example, one competitor has stated a variance of 2C in ambient temperatures while testing - such a high level of variance can significantly skew results.

Can it be a faulty products you got for the review?

That's a possibility, but I did test three of them - the 360mm, the 280mm, and the 240mm. If all three were defective, Arctic has QA problems.

And then you have contact pressure with this new contact frame. Done pressure tests? Thanks

I haven't done any pressure tests, but Gamers Nexus does.

Every test (e.g. from Gamers Nexus) shows that version III is better than II

I'm sure Liquid Freezer III is better than the old version. Both fail the current standards I judge coolers by.

(which you gave 4.5 stars).

Since then, I've changed my testing methodology and how I grade coolers. I'm always trying to figure out how to make my reviews better and more useful. One of the main reasons I choose my current platform was finding that Arctic's Liquid Freezer II failed to keep my system's i7-13700K under TJMax in maximum strength workloads, whereas DeepCool's LS720 was capable of doing so.

 

[tvargek]

interestingly, some sites say it's good and some it's bad, guess u can't trust no one any more...

 

[Albert.Thomas]

interestingly, some sites say it's good and some it's bad, guess u can't trust no one any more...

What I've realized from this is that perhaps I need to explain in more detail my testing methodology.

The way I test coolers on my latest platform was specifically chosen to represent the experience a user might actually experience in a real case and not on an open test bench which dramatically lowers cooling difficulty. Ambient temperatures are strictly controlled at 23C to prevent variance from impacting results. I could go on, but I'll save this for my upcoming review of Arctic's 36 air cooler.

 

[35below0]

Looking forward to Arctic 36, actually.

 

[MoxNix]

My understanding is a custom contact frame was used. Maybe that's the problem. Did the author even try testing the Freezer III with the contact frame it comes with instead of the one he likes?

 

[thestryker]

My understanding is a custom contact frame was used. Maybe that's the problem. Did the author even try testing the Freezer III with the contact frame it comes with instead of the one he likes?

You can't mount the Freezer III on Intel without using the contact frame that comes with it.

 

[Zenkai_1]

@

What I've realized from this is that perhaps I need to explain in more detail my testing methodology.

The way I test coolers on my latest platform was specifically chosen to represent the experience a user might actually experience in a real case and not on an open test bench which dramatically lowers cooling difficulty. Ambient temperatures are strictly controlled at 23C to prevent variance from impacting results. I could go on, but I'll save this for my upcoming review of Arctic's 36 air cooler.

Its not your testing, you are correct. I found this video and it confirms your findings are true. The LF3 is just generally meh, and underperforms especially on intel. Check this video I found. Later in the video even when they put high rpm server fans on all the other aio's the performance was still among the worst. Really disappointed. They are better options available.

View: https://youtu.be/pope6DUBjVA?t=508

 

[Albert.Thomas]

Its not your testing, you are correct.

I've also seen a few other reviews which align with my results lately, most recently from STS

View: https://youtu.be/YjoJoPxHClI?t=456


You'll see his results show the Iceberg Thermal IceFLOE Oasis 240 - which I previously reviewed here - performing less than 2C away from Arctic's Liquid Freezer 420.

But this is definitely about how people are testing. You absolutely do not see a bunch of reviewers saying that this is the 2nd coming of liquid cooling, and others saying that it is underwhelming, without some major flaws in testing methods.

I suspect that the reasons you're seeing other reviewers show more favorable results is down to the following possibilities:

• Lack of variable controls - i.e. results that have substantial variance in ambient temperature
• Testing on an open test bench - this doesn't accurately represent the conditions a CPU cooler would actually deal with for most users
• Testing on a thermal heatplate - this doesn't accurately represent conditions because the heat is evenly spread, whereas a CPUs largest sources of heat come from a few thermally dense hotspots.
• Testing on older, less thermally intensive CPUs. I only test coolers on Intel Raptor Lake or AMD Ryzen 7000 CPUs for a reason : older CPUs just aren't as difficult to cool due to a combination of lower clock speeds and less dense manufacturing processes.

 

[ViperThrall]

I don't understand this since I own and run an Artic Freezer III 360 ARGB? Not saying the numbers are not correct for an Intel processor but my 7950x at 179 watts runs at 5.7 MHz on 7 cores while the rest run at 5.4 under load? The price was stupid at $97 so I jumped on this cooler.
I have the headers split out to pump, VRM and fans. Everything is set via motherboard. Extremely quiet under load. The highest temps I'm recording are between 70C and 84C in my be Quiet 800 case with 4 140mm fans.
This processor is designed to go to 95C and doesn't even get close. Yes I could possibly push the voltages up but I am happy with the processor speed and not too crazy voltage hike.
Currently when under load the 140 mm case fans spin up to about 65% while the pump is at 100%, the VRM is at 90% and the fans on the Artic Freezer are at 82-87%. It is noticeable but for what the machine is doing still quiet in my small den/office space. I just need to keep the door open since it does heat this space up
I did return my Thermalright Peerless Assassin since it could not keep up with this setup. Under load the processor would hit 95C pretty quickly and the fans on the Assassin would really spin up hard.
For AMD fans out there the price to performance is kind of crazy if the sale price can still be had.
After tax and shipping the Peerless Assassin ARGB was almost $43 while at $97 I considered the Freezer III to be a hell of a buy.
Really like this AIO. Quiet, inexpensive, efficient and it looks good. High end Intel users probably need to stay away.:<

 

[thestryker]

I don't understand this since I own and run an Artic Freezer III 360 ARGB? Not saying the numbers are not correct for an Intel processor but my 7950x at 179 watts runs at 5.7 MHz on 7 cores while the rest run at 5.4 under load? The price was stupid at $97 so I jumped on this cooler.

AM5 != LGA1700

It's really as simple as that with the LFIII. If you check other reviews who do cover AM5 it's one of the best, if not the best in every size class.

 

[beyondlogic]

AM5 != LGA1700

It's really as simple as that with the LFIII. If you check other reviews who do cover AM5 it's one of the best, if not the best in every size class.

simple reason it runs like a trash keep on lga contact frame that intel slaps on these things is utter garbage

lga processors bend or bow with contact frame thus coolers dont make great contact with the die

the lever system doesn't work well because of the length of the processor on intel. works fine for amd cause its bloody square.

View: https://youtu.be/iYU1OskbY-Q?t=597

View: https://youtu.be/tSzKW53FQt8?t=112

 

[beyondlogic]

I don't understand this since I own and run an Artic Freezer III 360 ARGB? Not saying the numbers are not correct for an Intel processor but my 7950x at 179 watts runs at 5.7 MHz on 7 cores while the rest run at 5.4 under load? The price was stupid at $97 so I jumped on this cooler.
I have the headers split out to pump, VRM and fans. Everything is set via motherboard. Extremely quiet under load. The highest temps I'm recording are between 70C and 84C in my be Quiet 800 case with 4 140mm fans.
This processor is designed to go to 95C and doesn't even get close. Yes I could possibly push the voltages up but I am happy with the processor speed and not too crazy voltage hike.
Currently when under load the 140 mm case fans spin up to about 65% while the pump is at 100%, the VRM is at 90% and the fans on the Artic Freezer are at 82-87%. It is noticeable but for what the machine is doing still quiet in my small den/office space. I just need to keep the door open since it does heat this space up
I did return my Thermalright Peerless Assassin since it could not keep up with this setup. Under load the processor would hit 95C pretty quickly and the fans on the Assassin would really spin up hard.
For AMD fans out there the price to performance is kind of crazy if the sale price can still be had.
After tax and shipping the Peerless Assassin ARGB was almost $43 while at $97 I considered the Freezer III to be a hell of a buy.
Really like this AIO. Quiet, inexpensive, efficient and it looks good. High end Intel users probably need to stay away.:<

i would be interested in a test with thermalrights contact frame on intel and see if they issues with its performance vanish its probly a pressure issue

 

[thestryker]

simple reason it runs like a trash keep on lga bracket that intel slaps on these things is utter garbage

lga processors bend or bow with standard brackets thus coolers dont make great contact with the die

the lever system doesn't work well because of the length of the processor on intel. works fine for amd cause its bloody square.

View: https://youtu.be/iYU1OskbY-Q?t=597

View: https://youtu.be/tSzKW53FQt8?t=112

It's not this the LFIII LGA1700 mounting system is their own contact frame. GN's test indicated it doesn't make very good contact.

 

[beyondlogic]

It's not this the LFIII LGA1700 mounting system is their own contact frame. GN's test indicated it doesn't make very good contact.

that was my point im not saying its the lfIII coolers mounting system at fault. but the test in the video still has the crappy intel lga contact bracket.

so when you mount a water cooler on its going to perform worse.

lga stock bracket + IfIII bracket for intel = bad cooling performance why its probly because the cpu isnt making contact properly.

now if they try the test with a thermalright contact bracket and the IfIII i bet they would see better numbers.

( sorry for confusion lol) to me its just a bracket.


https://preview.redd.it/same-cpu-bu...bp&s=8cdd092fc90c35730222f9a0c9fa6b9bc7f4f71f

 

[thestryker]

that was my point im not saying its the lfIII coolers mounting system at fault. but the test in the video still has the crappy intel lga contact bracket.

so when you mount a water cooler on its going to perform worse.

lga stock bracket + IfIII bracket for intel = bad cooling performance why its probly because the cpu isnt making contact properly.

now if they try the test with a thermalright contact bracket and the IfIII i bet they would see better numbers.

( sorry for confusion lol) to me its just a bracket.


https://preview.redd.it/same-cpu-bu...bp&s=8cdd092fc90c35730222f9a0c9fa6b9bc7f4f71f

To install the LFIII you have to remove the Intel stock ILM because Arctic made their own contact frame. GN did a pressure test of this and found it wasn't all that good. You cannot install a contact frame and then install the LFIII.

 

[beyondlogic]

To install the LFIII you have to remove the Intel stock ILM because Arctic made their own contact frame. GN did a pressure test of this and found it wasn't all that good. You cannot install a contact frame and then install the LFIII.

well thats annoying i wonder if its the lfIII materials for the frame looks like plastic or rubber possible the material isnt ridged enough ? or expanding with the heat ?.