Review: MSI Core Frozr L processor cooler

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Nov 1, 2016.

  1. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    48,392
    Likes Received:
    18,560
    GPU:
    AMD | NVIDIA
    We test and review the MSI Core Frozr L processor cooler. MSI is placing the cooler in the market, likely made with the same designers and fab that manufacturers their TwiNFrozr GPU cooler. The end re...

    Review: MSI Core Frozr L processor cooler
     
  2. JonasBeckman

    JonasBeckman Ancient Guru

    Messages:
    17,564
    Likes Received:
    2,961
    GPU:
    XFX 7900XTX M'310
    Impressive results, just a bit shy of the older Noctua DH14 model which is a fair bit bulkier and heavier, not bad at all.

    EDIT: Noctua DH14 ~1200g with both fans whereas this according to specs weighs 960g, still heavy but a good reduction while offering similar performance.
    (Plus the DH14 uses two fans whereas this uses one but if I remember reviews right there's not a huge difference if you remove one of the fans from the DH14.)

    Can't really tell due to this motherboard in the review having a shroud or something over the CPU slot but I guess this fan isn't quite as restrictive with RAM size either depending on motherboard layout.

    EDIT: Falls behind a little bit more in the overclock test but not by that much, 3 degrees or so in idle and then pretty much even in the full load test which is the more important one.

    Though I guess air cooling caps out depending on various factors from idle temps to airflow and such whereas water cooling while a bit more work to set up has less restrictions.
    (And then there's oil cooling or more like submersion but I haven't seen that in a few years now so I guess it was too much work for only marginal improvements over water?)

    EDIT: Seems to be fairly silent too.

    Wonder how that 5960K 8-core CPU would fare? Guessing it's a bit above the 4790 here in terms of heat although if one can afford a 1000$+ dollar CPU just like that one could probably afford decent cooling for it heh not exactly standard hardware with such pricing though then again for "just" gaming a eight core CPU might be a bit overkill in most titles due to CPU scaling, "lower" end model but OC'd a bit to make up for it might fare better too due to that, probably getting a higher stable OC too.

    (Not that I would know but I assume with hardware like that or GPU's like the TitanX Pascal well the manufacturer might be a bit less willing to let the review website or such keep said hardware too so probably not the easiest to keep around for testing.)
     
    Last edited: Nov 1, 2016
  3. Koniakki

    Koniakki Guest

    Messages:
    2,843
    Likes Received:
    452
    GPU:
    ZOTAC GTX 1080Ti FE
    Looks like MSI has a winner here!

    Silent, performing great and good price! Well done! :thumbup:
     
  4. Han2K

    Han2K Master Guru

    Messages:
    306
    Likes Received:
    6
    GPU:
    MSI GTX1080 GX
    I wonder if the Corsair H110 and above are in Quiet or Performance mode in the temperature tests. I guess they are in Perf. mode cause they are slightly ahead of EK Predators.

    Apart from that it would have been very interesting to test the HS reviewed with the addition of a 2nd fan of similar characteristic in pull mode.
     

  5. Grafdelwer

    Grafdelwer Member

    Messages:
    14
    Likes Received:
    0
    GPU:
    MSI GTX 970 OC
    I wonder what the effect will be on the GPU's temperature if you take into consideration the direction of airflow. According to the images, the exhaust blows directly on the graphics card...just a thought.
     
  6. ezodagrom

    ezodagrom Guest

    Messages:
    599
    Likes Received:
    47
    GPU:
    MSI GTX 1660 Ti 6GB
  7. Koniakki

    Koniakki Guest

    Messages:
    2,843
    Likes Received:
    452
    GPU:
    ZOTAC GTX 1080Ti FE
    The reason for the specific orientation in the review was because HH used the Sabertooth Z97. It will not affect the majority of other motherboards.

    So sideway installation to avoid any thermal issues for the gpu will not be an issue..

    Just saying. :p

    Quote:

    "One problem you could run into is the base-plate mounting. It is a little big and if you have a motherboard (like we used) with thick plastic shielding and stuff, you might need to mount the cooler 90 degrees differently (as shown in the top photo). 98% of the motherboards do not have this issue though, we use the Z97 SaberTooth from ASUS for this very specific reason as it has shielding all over the place."
     
  8. Undying

    Undying Ancient Guru

    Messages:
    25,335
    Likes Received:
    12,746
    GPU:
    XFX RX6800XT 16GB
  9. sykozis

    sykozis Ancient Guru

    Messages:
    22,492
    Likes Received:
    1,537
    GPU:
    Asus RX6700XT
    So, where can I get one???
     
  10. Grafdelwer

    Grafdelwer Member

    Messages:
    14
    Likes Received:
    0
    GPU:
    MSI GTX 970 OC
    Ah, I see - missed that bit. Thanks for pointing that out.
     

  11. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    11,629
    Likes Received:
    1,119
    GPU:
    4090 FE H20
    Problem here is that since the 6700k uses TIM instead of solder and the fact that it has a tiny die, TIM cannot conduct heat fast enough..

    That is why all of the HSF coolers are so similiar.

    a 5960x(or 6 core) would be a better CPU to test with; it would show a coolers true potential and the DH-14 should gain another 5-10c advantage over this MSI cooler.
     
  12. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    48,392
    Likes Received:
    18,560
    GPU:
    AMD | NVIDIA
    You know, you are right and you have is a totally sound thesis, however people want to compare coolers with processors they can afford / own and not compare it towards processors only 0.1% of the PC community have.

    We did that in the past and people seriously hated the usage of an enthusiast class processor for cooling reviews. For the test itself objectively you are right, it's just not what the majority of our reader-base would like to read about. They want to be able to compare it to a setup they own and can afford (or is reasonably comparative to that).

    It actually is a very similar situation with our graphics card reviews, the reviews often get attacked due to the fact I use an 8-core 5960X @ 4.3 GHz where I often get the remark that hardly anyone owns /uses such a processor. See I am testing the full capacity of a graphics card. and am not trying to limit it. The huge majority of the reader-base is using a quad-core processor. The number of times that people claimed that our 3DMark results are screwed due to using an 8-core part I cannot even count anymore. Neither can I convince the reader-base that is it better to use a massive processor for testing. e.g., I use such a high-end tweaked CPU not limit (if possible) the higher-end and enthusiast class graphics cards. E.g. I do not want to CPU bottleneck a test environment at lower resolutions. Two years after using this processor people finally and slowly are accepting the 8-core processor as a valid component in the test-setup,

    I purchased a 6950X 10-core processor for GFX reviews, yet cannot use in my graphics card reviews for exactly the reasoning I just described, people would spit on the reviews if they see a 10-core beast at a 4.3 GHz clock per core. But considering it is a 1200 USD part, people absolutely and irrefutable refuse to accept it as a proper component in a test environment. It is all for the very same reason, people want to compare with what they have at home (within reason and acceptable margin).

    It is the same with CPU cooler reviews ... we need to stay a little in the affordable range. This is why I introduced the overclocking segment in the cooler reviews .. if the cooler can take it we'll take the processor up and over 1.40 Volts. The high voltage OC results as far as I am concerned are a definitive enough answer to your question, as the tweak really stresses the cooler.

    Okay, that's way more words then I wanted to type in this reply lol :nerd:
     
  13. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    11,629
    Likes Received:
    1,119
    GPU:
    4090 FE H20
    Yes I understand where you are getting at.

    I don't get how people don't understand when reviewing a GPU that you want it to be the bottleneck of the system.

    What good is testing it with an i5 and only getting 70% usage out of a titan X pascal?

    But anyways, I think you should use a 6 core as the CPU for thermal testing.
    a 5820k/6800k are very obtainable, they are only like $40 more expensive than a 6700k.

    I get that most users will have a CPU that uses TIM instead of solder but it skews the results for others.
    Someone will look at the review and see that this $30 cooler is only a couple degrees from a $100 cooler and think that $100 cooler is a waste of money.
    Then they proceed to use a $30 cooler on their 6/8 core and wonder why temperatures are so high under load.

    That's why IMO a 6 core would be a perfect medium to test thermals with.
    It gives a more realistic representation of a coolers potential.
     
  14. tsunami231

    tsunami231 Ancient Guru

    Messages:
    14,725
    Likes Received:
    1,854
    GPU:
    EVGA 1070Ti Black
    That lil cool has some nice performance it even gives the noctau D-14 a run for its money
     
  15. sykozis

    sykozis Ancient Guru

    Messages:
    22,492
    Likes Received:
    1,537
    GPU:
    Asus RX6700XT
    Considering your previous statement that the 6700K uses TIM and not solder, using the 4970K makes considerably more sense. The 4970K runs hot as hell. If the CPU cooler can cope with the heat produced by the 4970K, and keep that particular processor reasonably cool, it's safe to assume that it will do the same with a processor that has proper heat transfer. Of course, you are pointing out a limitation of Noctua's coolers....

    I'd expect a $100 cooler to perform just as well, in regards to other coolers, regardless of the thermal transfer media being used between the CPU core and IHS......but that's just how most of the market thinks.
     

  16. Texter

    Texter Guest

    Messages:
    3,275
    Likes Received:
    332
    GPU:
    Club3d GF6800GT 256MB AGP
    Great job from MSI...nice price too.
     
  17. Agent-A01

    Agent-A01 Ancient Guru

    Messages:
    11,629
    Likes Received:
    1,119
    GPU:
    4090 FE H20
    The chip isn't hot compared to a 6 or 8 core.
    4790k vs 6700k put out the same heat and they have the same limitation in which they are using TIM with poor conductivity(in range of 7-8 W/m-K).

    Delidding it and using a liquid metal TIM brings 20-30c drop in temperature.
    This would make results much more spread out too(i suppose this could be an alternative)

    The only reason the AIOs are beyond even the best air coolers are due to water's ability to conduct heat better.

    You compare a top end AIO(excluding ones like the ek 360rad cooler) to the best air cooler and you'll get similar results with a soldered CPU.
    Even a cheap AIO is better than a NH-D15 cooler coupled with 6700k(which isn't the case on a soldered CPU)

    It just doesn't make sense that the bottom-end air coolers are the same as the top end ones.

    Using a CPU with TIM instead of solder doesn't accurately depict the best or most realistic results.

    A 5820k would be the perfect medium.
    It's affordable, you can get a new one for $340. X99 is general doesn't cost much more anymore.

    Instead of having a 2c difference between a cheap cooler and a top end one the results would be much more spread out.

    Reviews like this(not pointing/blaming HH as many review sites do the same thing) make people think that all air coolers perform similar; i.e. it's not necessary to spend more than $40.

    Hell, AMD's Zen could be a killer CPU and force Intel to 'better' their product by using solder for all their CPUs.
    Then results like these would be invalidated.

    Just makes more sense now and the future to use a 5820k~.

    or include two separate systems for temperature results, but i know this is a lot of extra work. HH does stuff above and beyond already.

    That's my 2c.
     
    Last edited: Nov 1, 2016
  18. Stormyandcold

    Stormyandcold Ancient Guru

    Messages:
    5,872
    Likes Received:
    446
    GPU:
    RTX3080ti Founders
    I wish manufacturers would include another set of clips with all coolers as standard.
     
  19. ezodagrom

    ezodagrom Guest

    Messages:
    599
    Likes Received:
    47
    GPU:
    MSI GTX 1660 Ti 6GB
    Sorry for bumping this, but, any idea where or when this cooler will be available for purchase? I just can't find it anywhere (interested in it sometime next year once I get a new CPU).
     

Share This Page