Arctic Liquid Freezer III Pro Review: The best AIO for Ryzen 9950X3D, and Intel too

The Liquid Freezer III Pro loudly raises the bar for performance and value, handling the hottest CPUs on the market with ease.

Arctic Liquid Freezer III Pro
Editor's Choice
(Image credit: © Tom's Hardware)

Tom's Hardware Verdict

I couldn’t believe my eyes when I saw the stunning results of my testing of Arctic’s Liquid Freezer III. This cooler earns its five-star rating by combining chart-topping performance with a surprisingly reasonable price.

Pros

  • +

    The strongest AIO I’ve tested, bar none

  • +

    Chart-topping noise-normalized performance

  • +

    Easily handles even the 9950X3D with PBO enabled

  • +

    Six-year warranty

  • +

    Very reasonable MSRP of $125

  • +

    Up to 6.9 mm H2O static pressure!

Cons

  • -

    Intel users still have to install a custom contact frame

  • -

    This is not a quiet cooler, so you may want to limit the pump and fan speeds

Why you can trust Tom's Hardware Our expert reviewers spend hours testing and comparing products and services so you can choose the best for you. Find out more about how we test.

Arctic is a name well-known and beloved by many PC enthusiasts for providing high-value, quality cooling solutions. I caused a bit of a controversy with my last review, where I didn’t mince words, stating “the Liquid Freezer III is unimpressive.”

Well, it appears that Arctic took my feedback to heart. We’re looking at the Liquid Freezer III Pro today, and let me tell you guys, they take the pro moniker very seriously. If I didn’t know better, I’d swear they designed this cooler just for me. Its performance blew away my expectations – I honestly couldn’t believe my own results until I tested it with both Intel’s Core i7-14700K and AMD’s Ryzen 9 9950X3D CPUs.

Latest Videos FromTom's Hardware
Albert Thomas
Freelancer, CPU Cooling Reviewer

Albert Thomas is a contributor for Tom’s Hardware, primarily covering CPU cooling reviews.

  • cknobman
    Wow this thing is a beast.
    Reply
  • JayGau
    It's astonishing to give 5 stars to an AIO that doesn't have the most basic feature any AIOs should have: a liquid temperature sensor. It's like buying a car without a temperature gauge. The liquid temperature is the most important and critical data on your AIO. The fans are not cooling down the CPU, they are cooling down the liquid.

    You say the fans are loud, and so many people are complaining that AIOs are louder than air coolers, which for me is a total nonsense. AIOs are much quieter than air coolers, it's not even a contest, as long as you use them properly. Of course if you set the fans with the CPU temp you gonna have 3 fans (for a 360mm) bursting like hell every time the CPU does something. But if you set them with the liquid temperature instead (which is what the fans are cooling down), you barely notice them since the liquid never goes really hot and its temperature changes very slowly.

    It's unbelievable that in 2025 most AIOs don't even have this functionality. But I guess manufacturers don't really care since they know most people don't even know how those coolers work anyway.
    Reply
  • Elusive Ruse
    Thanks, these results are more in line with what I expect from an Arctic AIO, I will be on my LFII 360 as long as I keep my 7950X because it’s been incredible but my next AIO is most definitely going to be this one.
    Reply
  • Albert.Thomas
    JayGau said:
    It's astonishing to give 5 stars to an AIO that doesn't have the most basic feature any AIOs should have: a liquid temperature sensor. It's like buying a car without a temperature gauge. The liquid temperature is the most important and critical data on your AIO. The fans are not cooling down the CPU, they are cooling down the liquid.
    I understand why some people would prefer to tie fan speeds to liquid temperatures, but I don't like it *most* of the time.

    While I appreciate the lack of fan bursts, there's a few reasons I do not use coolers this way.

    Scenario #1: Near-term throttling. When you use a cooler that operates based on liquid temp, your CPU can hit TJMax and throttle as a result - while the cooler stays in low operation because the liquid temperature hasn't warmed up yet.

    Scenario #2: Extended high speed fan duration. If you run a high intensity load for more than a few moments - lets say a full CPU load for 10 minutes - it is gonna take a while for the liquid temp to cool. A traditional cooler will lower fan speeds almost immediately after a workload ends. However, a cooler tied to liquid temperature will take two or three minutes to return to low fan speeds.

    JayGau said:
    You say the fans are loud, and so many people are complaining that AIOs are louder than air coolers, which for me is a total nonsense. AIOs are much quieter than air coolers, it's not even a contest, as long as you use them properly. Of course if you set the fans with the CPU temp you gonna have 3 fans (for a 360mm) bursting like hell every time the CPU does something. But if you set them with the liquid temperature instead (which is what the fans are cooling down), you barely notice them since the liquid never goes really hot and its temperature changes very slowly.
    I test things using the motherboard's default fan curve. While you may or may not agree with this, this is how most users will run their coolers.

    Noise normalized results are included to show how you can still have higher performance while operating quietly.
    Reply
  • LookHereBub
    Interesting. After much research I purchased the non-pro LF3 middle of last year when I built my current system. I have found it excellent in all respects, keeping my 12900ks from throttling even under heavy assault using cinebench. Under normal use with most games CPU rarely gets over 50-60C. Not sure I would need to upgrade, but nice to see them making good improvements. Nice review, thanks.
    Reply
  • AmirTHforum
    This review is very disappointing!
    I just built a new computer, and I made my decision on your review on April 6, 2024.

    https://www.tomshardware.com/pc-components/liquid-cooling/arctic-liquid-freezer-iii-aio-review
    When I saw this review I said to myself; ok, maybe someone else at TomsHardware has a different opinion; but no it's the same person doing the same review on the same product. Now all of a sudden, you go from 2.5 stars to 5 stars.
    Now I can understand rerunning the test and using "Updated cooler testing methodology" (August 18, 2024).

    https://www.tomshardware.com/pc-components/liquid-cooling/a-second-look-at-arctics-liquid-freezer-iii-using-our-updated-cooler-testing-methodology#xenforo-comments-3852191
    But how do you go from; "Unimpressive" to "The best AIO for Ryzen 9950X3D, and Intel too"?

    What happened?
    Did Arctic call you and yell at you for giving a bad (but true) review?
    Or is Arctic Liquid Freezer 3 is a great product and your initial testing was that wrong?

    One last thing; you (Albert Thomas) got a little defensive when readers questioned your review (April 6, 2024), and you were adamant, that your methodology was better and more accurate that the rest of the reviewers online.
    Reply
  • -Fran-
    Thanks for the data as always Albert.

    Would you ever test the lil' bro of this one: LF3-240, or at least the 280, if there's a version of them?

    Also, are you sure it's 200W for the 9950X3D? Mine goes all the way to ~240W at stock. Since my LF2-240 can't keep up with it past ~200W (I may need to consider de-lidding xD), it only bursts until temps catch up and it ramps down.

    Regards.
    Reply
  • Albert.Thomas
    AmirTHforum said:
    When I saw this review I said to myself; ok, maybe someone else at TomsHardware has a different opinion; but no it's the same person doing the same review on the same product. Now all of a sudden, you go from 2.5 stars to 5 stars.
    Please re-read the review in detail.

    This is NOT the same product. If you pay attention to the results using Intel's i7-14700K, you'll see results with both the original Arctic Liquid Freezer III reviewed last year and the product reviewed today which is the Liquid Freezer III Pro
    Perhaps Arctic should have given this AIO a new branding to avoid any confusion
    Reply
  • Albert.Thomas
    -Fran- said:
    Thanks for the data as always Albert.

    Would you ever test the lil' bro of this one: LF3-240, or at least the 280, if there's a version of them?

    Also, are you sure it's 200W for the 9950X3D? Mine goes all the way to ~240W at stock. Since my LF2-240 can't keep up with it past ~200W (I may need to consider de-lidding xD), it only bursts until temps catch up and it ramps down.

    Regards.
    If Arctic releases this AIO in another size, I'm sure management wouldn't mind a look at another variant given this model's impressive performance.
    Reply
  • Elusive Ruse
    -Fran- said:
    Thanks for the data as always Albert.

    Would you ever test the lil' bro of this one: LF3-240, or at least the 280, if there's a version of them?

    Also, are you sure it's 200W for the 9950X3D? Mine goes all the way to ~240W at stock. Since my LF2-240 can't keep up with it past ~200W (I may need to consider de-lidding xD), it only bursts until temps catch up and it ramps down.

    Regards.
    I highly doubt it goes that high without PBO enabled.

    https://tpucdn.com/review/amd-ryzen-9-9950x3d/images/power-multithread.png
    Reply