MSI MAG CORELIQUID 360R Reader Review by TAKTAK - Dec 2020

Despite being set to public, images don't show unless you're logged in to the forum
I'll try and fix
Zak Edit - I got the boss to mend it personally :) images now public

Context
Uncle Zakky kindly arranged for MSI to ship a MAG CORELIQUID 360R over to me for a HEXUS user review.

I'll start with a disclosure notice
I have no vested interest in MSI as a company and I am not being 'paid' for my review of the MSI MAG CORELIQUID 360R, however, a transaction has occurred in that I have been provided with a review sample in return for review.
All views are my own and are in no way swayed/influenced by HEXUS and/or MSI, nor anyone else for that matter.

---

Quick links

• Pre-Delivery and Unboxing, this post.
• A closer Look and Fitting
• Testing
• Conclusions

---

Pre-delivery
Prior to receipt of the cooler, I figured that I'd take a peep at the MSI landing site for the AIO (here)

Some of the marketing copy isn't particularly great, with one example being the following:

https://forums.hexus.net/members/tak...-siteshot1.jpg

Portions of the gaming industry are, disappointingly, split across irrelevant emotional 'factions' (PC master-race, Team Green, Team Red, etc etc). What also comes with this territory is a desire to have the best and not be seen as inferior (yawn). To that point, in my opinion, MSI are immediately at risk of putting off enthusiast consumers by employing statements like 'making it the best choice for entry level gamers.'
"Oh, well I don't consider myself an entry level gamer, so this must not be for me. I'll go to another product which is the best."
I'm reasonably confident that MSI aren't trying to solely target a 360 AIO to 'entry level gamers', but there are many ways to read that statement...


MSI claims that the pump has been integrated into the radiator for sound and longevity purposes. I have no objective way of testing the veracity of that statement but I will point out that Asetek hold, and protect, the patent for block mounted pumps.

https://forums.hexus.net/members/tak...-siteshot2.jpg

Nevertheless, it is interesting to see that MSI have chosen to go against the tide and partner with a different OEM to Asetek. Consumer choice is always a good thing to have, and greater adoption of alternatives funds further R&D. Perhaps the Asetek designs have finally be gazumped?

Heading further down the site, we run in to an oddly orphaned web asset repurpose - The only reason I can see for the presence of the 'DIY 2.0' section is to upsell MSI motherboards and cases, only one of the clickable sections holds any direct relevance and realistically the section just gets in the way and creates confusion rather than adding anything of value to the product page.

https://forums.hexus.net/members/tak...-siteshot3.jpg

Based on MSI's recommendations I am expecting the MAG CORELIQUID 360R to perform very well on my 3700X, given that they recommend using the MAG CORELIQUID 240R for my CPU.
Claimed socket compatibility is great, and three 4-pin ARGB fans are included with a quoted noise level of 14.3 .. 34.3 dBA for an airflow of 21.63 .. 78.73 CFM.

N.B. Whilst I have looked over the MSI pages for the AIO, I haven't actively looked at other reviews of the AIO, in order to avoid any potential skew or regurgitation.

---

Post-delivery
Packaging
The top of the box is standard fare, showing a product still shot:

https://forums.hexus.net/members/tak...79-boxtop.jpeg

The bottom of the box is also par for the course:

https://forums.hexus.net/members/tak...-boxbottom.jpg

Internally, everything appears to be well protected with a decent level of impact protection and an additional thin cardboard sleeve around the radiator:

https://forums.hexus.net/members/tak...xinternals.jpg

Contents
Getting to the internals we have 3x aRGB fans and 1x AIO assembly:

https://forums.hexus.net/members/tak...incontents.jpg

In addition, we have various mounting accessories for different sockets (1366/1156/1155/1151/1150/AM4/AM3/TR4/sTRX4/SP3). Along with a fan header splitter cable and a resistive dropper cable (for reducing the fan speed):

https://forums.hexus.net/members/tak...ntingequip.jpg

I have some reservations about the Threadripper support, whilst there is a supplied bracket to mount the cooler, a bracket doesn't magically increase the size of the CPU block and Threadripper parts are physically larger.

A slightly closer look and fitting
The sides of the radiator have a subtle MSI logo, just gloss black on matt black:

https://forums.hexus.net/members/tak...ideaccent.jpeg

The CPU block has a dragon on it (it's MSI, it would frankly be strange if it didn't):

https://forums.hexus.net/members/tak...ockgraphic.jpg

The fan corners all have a rubber damper installed, it's a small feature but I'm glad for it. I'm unsure how effective it will actually be in compression, but should prevent some potential rattling either way:

https://forums.hexus.net/members/tak...berdamper.jpeg

One thing to note with the block and fan RGB LEDs is that they're the 5V addressable type, so if your motherboard is the 12V kind then you're in the territory of requiring an extra bit of hardware in order to drive them.
As it turns out, I'm in this latter camp and have no way of plugging aRGB LEDs in (I didn't realise until planning out cable runs for the RGB fan cabling)

Thankfully, the block has provisioned for this and the aRGB cable is detachable:

https://forums.hexus.net/members/tak...ablecable.jpeg

More on cable routing - With the pump mounted mid radiator, the power cable comes out in an awkward place. MSI have mitigated this reasonably well by providing slots in the shroud on either side:

https://forums.hexus.net/members/tak...tingcutout.jpg

The presence of a cable slot on each side is likely down to reducing piece parts for manufacture, rather than trying to provide flexibility. If you were to use the slot on the other end of the rad, then it doesn't leave you a great deal of cable to play with, plus your cable would be running the full length of the rad surface.

Strangely the slot appears to be absent on the MSI marketing shots...

---

Previous setup
This is all inside a Corsair Obsidian 800D (there are threads on this forum of previous builds I have in this case, including a long since decommissioned custom water loop.)
My previous setup used the AMD Wraith Prism along with 3 Noctua NF-P12s in the top venting out.

https://forums.hexus.net/members/tak...viousbuild.jpg

N.B. I've only been using the Wraith Prism for a couple of months, it happened to come with an upgrade bundle from SCAN, I just hadn't got around to replacing it yet (I moved from an i7 860, remember those?!).

Build
If you're like me then you'll probably attempt to plan where your cable runs are heading, given that the MSI fans each have two cables coming out of them (one fan cable, the other being aRGB). It would be nice if manufacturers provided options for cable routing built in to product as most users route cables in a limited set of ways. In this case, I'm taking all cables to one end to use a single exit point to the rear of my case:

https://forums.hexus.net/members/tak...lerouting.jpeg

The pump cable coming out at the rad presents another issue - In my experience, most pump headers are located right next to the CPU socket, meaning that you need to run a cable a hefty way into the board area :(
Because of cable lengths and the sheer size of my case, I\'m forced into mounting the rad in one orientation, else the pump cable wouldn\'t reach without an extender.

https://forums.hexus.net/members/tak...-postbuild.jpg
N.B. When removing the Noctua fans, I also removed my fan controller and connected the 3x case fans directly to the motherboard. As a result the fan speed currently ramps when PBO kicks in. I need to change how fans are controlled, but this doesn't affect the setup.

Testing

Test Platform

• CPU - Ryzen 3700X
• Motherboard - ASUS ROG STRIX B450-F GAMING
• RAM - 16GB Corsair Vengeance LPX 3200MHz DDR4
• GPU - 8GB ASUS Radeon RX 5700 XT 50th Anniversary
• Case - Corsair Obsidian 800D
• PSU - EVGA 650GQ
• SSD - SAMSUNG 500GB 970 EVO
• OS - Windows 10 (2004 19041.685) [N.B. There were some Windows updates between the Wraith Prism and MSI AIO tests]
• Radeon Driver - 20.11.2
• AMD Chipset Driver - 2.07.14.327

Thoughts on test approach
As this is a CPU cooler, I've largely focussed on isolated CPU thermal loads using Prime95.
In order to gauge differences between average and intense load performance, I used Prime95's Blend test for mixed FFTs.

Further to the focussed CPU torture testing, I threw in a couple of game benchmarks in order to highlight any changes to potential bottlenecking relating to thermal throttling of boost clocks.

Core boost clocks and thermals were dumped to CSV using HWInfo and graphed using Google Sheets (raw data)

---

I must prefix this test data by pointing out that I dun goofed.

My motherboard was auto overclocking my CPU and under the Wraith Prism my motherboard was setting clocks to 3.8GHz and disabling PBO, I was enabling PBO and resetting to defaults using Ryzen Master on boot as I couldn't initially work out why my PBO wasn't working on boot.

Installing the MSI AIO, my motherboard clocked to 3.9GHz (which was the red flag I needed to realise that something was meddling). I reset my BIOS and then set memory back to the correct profile, but left CPU at the stock 3.6GHz with PBO enabled, no longer having to set it in Ryzen Master on boot.

I intend to expand the test set by re-doing the Wraith tests under identical settings. To do that, I need to order some thermal paste, I'll then re-test the MSI AIO with different thermal paste too.

end of goof

Goof being said, I don't think it actually affects the results as during testing the clocks are sat on boost for the majority of the time.

Prime95 - Blend torture test (PBO Enabled)
AMD Wraith Prism

https://forums.hexus.net/members/tak...bo-enabled.jpg

N.B. the temperature rise ~7min is due to a change in the type of FFT being used by Prime95 during the blend test. Data prior to the 7min mark is reasonably indicative of temperatures I see whilst playing games.
N.B.B. I terminated the run when temperatures reached 93c; One can reasonably assume that it would keep rising, especially if the GPU was also under load transferring heat to the same environment.

MAG CORELIQUID 360R

https://forums.hexus.net/members/tak...ture4993-a.png

As shown above, the maximum temperature is dropped by +20c.

Although for me, the most interesting change is ~17.5 minutes where the temperature drops rapidly when the load is removed. This highlights that we have potential to reduce temperatures further if we can move heat away from the CPU quicker by improving the interface. The overall system is far below that of the CPU temperature.

Boost clocks are generally higher and more stable across all cores.

---

Cinebench 20
AMD Wraith Prism

https://forums.hexus.net/members/tak...r-ryzenpbo.PNG
N.B. Here you can see the 3700x listed as 3800MHz (instead of 3600MHz) due to the goof.
N.B.B. The other result you can see on here is a run with PBO disabled

MAG CORELIQUID 360R

https://forums.hexus.net/members/tak...r-ryzenpbo.PNG

Despite the goof, the MSI AIO comes out with a higher score, I believe this is due to sitting on boost clocks during the whole test, making the minor overclock on the Wraith Prism test irrelevant.

---

Age of Empires 2: Definitive Edition
AMD Wraith Prism

https://forums.hexus.net/members/tak...wraithaoe2.jpg

MAG CORELIQUID 360R

https://forums.hexus.net/members/tak...97-msiaoe2.jpg

Again, the MSI AIO provides a slight performance increase due to it's ability to provide more stable and higher boost clocks.

---

Borderlands 3
AMD Wraith Prism

- FramesPerSecondAvg: 71.95
- FrameTimeMsAvg: 13.90

MAG CORELIQUID 360R

- FramesPerSecondAvg: 73.55
- FrameTimeMsAvg: 13.60

Again, showing a minor improvement for the MSI AIO test BUT this is also right down near the noise floor, so could just be natural variation between tests.

Conclusions
Noise-wise the pump produces a low whirring noise, in my setup that noise is lower than that of other case fans. In silent set-ups it could prove to be an issue, but that's no different to any other AIO beyond the noise source now being in the radiator rather than the CPU block.

The Ryzen processors really do benefit from improved cooling, with performance upticks across the board due to increased and more stable boost clocks. I think there is still some potential additional performance to be had in improving the CPU interface to the cooler (whether than be better thermal paste or lapping the IHS).

The styling is clean and there is at least some attention to detail with detachable block RGB cable and different options for routing the pump power depending on your case setup. Plus you can orient the dragon however you like.

The MSI marketing copy could do with being revisited with thought put into who they're trying to market to and how people are going to be reading the marketing materials.

Pros:

• Core clock performance boost on Ryzen parts isn't to be sniffed at,
• Non-Asetek OEM = More-competition = More-better,
• Can be used without introducing RGB to your build.

Cons:

• Pump mounted in radiator - MSI claims that it helps with noise, but any pump noise is now being produced right at the top of my case, closer to my ears, rather than in the depths of my system...
• If you have the wrong RGB standard then no shiny shiny for you.

Other:

• Threadripper support is nice to see, but I'm dubious how effective it would actually be at contacting the whole heat area.

---

N.B. I'm unable to draw any conclusions about the performance within its class (i.e. comparisons to other 360 AIOs at a similar price point), as such I'd highly recommend looking over other reviews for that information.

What about the sound? How noisy is it?

Quote:

---

Originally Posted by Jonj1611

What about the sound? How noisy is it?

---

Subjectively, on par with other AIOs I've experienced - No worse, no better - Louder than a well designed custom loop, but also nowhere near the cost.
Comparatively to the Wraith Prism, leaps and bounds better.

It is significantly quieter than my 5700XT which dominates everything else when playing games. I still need to do some tweaking to fan profiles to prevent the PBO temperature spikes causing fan boosting in and out (which is quite frankly, annoying).

Edit: Fan boosting fixed by changing thermal source in BIOS

So the big question then..... Would you spend your own hard earned cash on one?

I'm in the process of putting together a 5800X build. That doesn't come with a cooler, so I'm trying to decide. I've settled pretty much on RAM and Mobo and it's going in a Phanteks Evolv X case (which I already have) so plenty of room.

Quote:

---

Originally Posted by b0redom

So the big question then..... Would you spend your own hard earned cash on one?

I'm in the process of putting together a 5800X build. That doesn't come with a cooler, so I'm trying to decide. I've settled pretty much on RAM and Mobo and it's going in a Phanteks Evolv X case (which I already have) so plenty of room.

---

I think I'd definitely get an AIO, but I also think that I'd angle towards a CPU block mounted pump (for better cable routing if nothing else). I have no need for RGB (individually addressable or not) so those aspects are wasted on me. I was intending on setting RGB up and seeing what it was like (I had it running on the Wraith Prism) but hardware constraints forced a stealth install.

I don't get the feeling that the MSI AIO is anything other than run-of-the-mill BUT I haven't been able to test performance against other 360 AIOs to form an objective viewpoint. I wouldn't discount it, but it also wouldn't be the automatic choice. Ultimately, comparative performance in the same class would seal the deal :)

For reference, before this came up I was mulling over a Noctua NH-D15 and also waiting on Lian Li's Galahad reviews. I don't recall ever seeing MSI in an AIO comparative roundup.
I stopped considering the Noctua due to potential future size interference issues, whereas I can shoe-horn a triple rad in many places.

Quote:

---

Originally Posted by TAKTAK

I think I'd definitely get an AIO, but I also think that I'd angle towards a CPU block mounted pump (for better cable routing if nothing else). I have no need for RGB (individually addressable or not) so those aspects are wasted on me. I was intending on setting RGB up and seeing what it was like (I had it running on the Wraith Prism) but hardware constraints forced a stealth install.

I don't get the feeling that the MSI AIO is anything other than run-of-the-mill BUT I haven't been able to test performance against other 360 AIOs to form an objective viewpoint. I wouldn't discount it, but it also wouldn't be the automatic choice. Ultimately, comparative performance in the same class would seal the deal :)

For reference, before this came up I was mulling over a Noctua NH-D15 and also waiting on Lian Li's Galahad reviews. I don't recall ever seeing MSI in an AIO comparative roundup.
I stopped considering the Noctua due to potential future size interference issues, whereas I can shoe-horn a triple rad in many places.

---

the Noctua NH-U14S may be a better bet. It sits on AM4 clear of the RAM with 2 fans on it so you can change RAM without even having to remove a fan, nevermind the cooler itself. I get decent temps with a 3900x in there. Damn I wish I had been able to take part and compare this thing to it.

Interestingly the unit I've been sent doesn't have a cable channel both sides for the pump wire. I even went back to check after reading this and it's definitely only on one side.

Different batch/revision possibly? Mine has serial 3602008154763

Quote:

---

Originally Posted by spacein_vader

Interestingly the unit I've been sent doesn't have a cable channel both sides for the pump wire. I even went back to check after reading this and it's definitely only on one side.

Different batch/revision possibly? Mine has serial 3602008154763

---

That's intriguing - I definitely have one each side at opposite ends (I did just have to check though!)

Serial format looks different too: 3067ZW1C3181310027007035

UPC: 824142205686
UCC: 10824142205683
EAN: 4719072697570

Quote:

---

Originally Posted by spacein_vader

Interestingly the unit I've been sent doesn't have a cable channel both sides for the pump wire. I even went back to check after reading this and it's definitely only on one side.

Different batch/revision possibly? Mine has serial 3602008154763

---

Same here - at least I didn't spot it if it had one...i was looking for it as it would have been handy

Quote:

---

Originally Posted by TAKTAK

That's intriguing - I definitely have one each side at opposite ends (I did just have to check though!)

Serial format looks different too: 3067ZW1C3181310027007035

---

Ah, I was working off the label on the rad. The box code is similar to yours but different: 3067ZW1C3181310027002898

What a superb job!!!

TAKTAK...you've been a super star! :)
That's a fantastic set of images and data :)

Thank you

Quote:

---

Originally Posted by Spud1

Same here - at least I didn't spot it if it had one...i was looking for it as it would have been handy

---

The plot thickens!

Quote:

---

Originally Posted by spacein_vader

Ah, I was working off the label on the rad. The box code is similar to yours but different: 3067ZW1C3181310027002898

---

My rad label and box S/N both match, I just included the other data from the box in-case any of the rest differed :)

Quote:

---

Originally Posted by Zak33

What a superb job!!!

TAKTAK...you've been a super star! :)
That's a fantastic set of images and data :)

Thank you

---

:hexlub: :hexlub: :hexlub: :hexlub:

Is that on AM4?

Did you have any trouble getting the mounting clip with the screw on? I've just installed a Celsius s36 and it was a nightmare trying to get the clip on. Felt like I was going to break something.

Quote:

---

Originally Posted by virtuo

Is that on AM4?

Did you have any trouble getting the mounting clip with the screw on? I've just installed a Celsius s36 and it was a nightmare trying to get the clip on. Felt like I was going to break something.

---

Yep, on AM4, mounting was easy for this (GPU still in-situ).
I loosely attached the clips to the CPU block (two piece per side, 1 clip with threaded bar and then a nut with a captive spring), then simply positioned and tightened 🙂

In hindsight, I should have detailed the mounting mechanism in the post; thanks for asking the question!