MSI MEG X299 CREATION

[Black Haru]

From the elite family of products dubbed MEG (MSI Enthusiast Gaming) comes another cutting-edge board with unique features and accessories. The MEG X299 CREATION is yet another board that promises to go above and beyond, so I took the review and turned it to 11.

Show full review

 

[diatribe]

Did you rip off the knob after you turned it to 11?

 

[Space Lynx]

waste of money since pci 4.0 is literally less than 5 months out. and samsung pci 4.0 NVME prob soon after.

 

[Dammeron]

 

[Rivage]

For US$599.99 it has no backplate.

 

[Black Haru]

For US$599.99 it has no backplate.

What is the point of a backplate on a motherboard?

 

[John Naylor]

Same with anything else ... that big heavy plate is a nice heat sink.

 

[Black Haru]

Same with anything else ... that big heavy plate is a nice heat sink.

I'm not seeing it. On the back of a motherboard there are very few components a backplate could interface with in a thermally meaningful way. These few include:

The VRM heat sink (via the mounting screws) - this is the most realistic, but using mounting screws to sink heat through the board is increadibly inefficient. You would be far better served by improving the VRM heat sink design.

The Chipset heat sink - the exact same as the VRM above, with the addition that the chipset really doesn't produce much heat anyway.

The back of the CPU socket - this is the most interesting possibility, but is pretty much infeasible. Because manufacturers have to account for a vast array of coolers and cooler mounting solutions, any backplate would really need to leave this area open.
Additionally, CPU socket hardware is produced by Lotes, not the motherboard maker. I haven't actually checked, but I'm confiden't that this is a requirement to meet Intel/AMD spec standards. So your backplate has to conform with the standard Lotes socket (achievable but inefficient).
Where this one might work is on boards with built in mono-blocks, since the block mounting solution is designed in-house. It would still be a pain, and of little to no benefit.

Also, the back of a motherboard gets zero or near zero air flow in any realistic build. A backplate has a real risk of trapping heat behind the board.

Where I see this concept maybe having merit is on ITX boards. To cram in as many features as possible high end SFF boards often put controllers or even M.2 slots on the back of the board. A traditional heat sink is not an option for clearance reasons, but a backplate could be a realistic alternative.

 

[Mescalamba]

Hm, maybe more thorough component dissection next time, VRM particularly?

Like I can make educated guess and say its 6 phases doubled, but is it? Who knows.. Definitely could use heatsink free shot.

x399 version has very nice overkill on VRM, but is x299 same or not? Yea that would be good to find out. Preferably WITHOUT buying one, which is why we have reviews.

Yea and 3x 8 pin is utterly useless. Just marketing BS.

But I agree on back plate. Thats another marketing BS. Has zero point, pretty much like those metal slabs on M.2 which are only good to keep heat there, not cool it.

 

[Black Haru]

Hm, maybe more thorough component dissection next time, VRM particularly?

Like I can make educated guess and say its 6 phases doubled, but is it? Who knows.. Definitely could use heatsink free shot.

x399 version has very nice overkill on VRM, but is x299 same or not? Yea that would be good to find out. Preferably WITHOUT buying one, which is why we have reviews.

Yea and 3x 8 pin is utterly useless. Just marketing BS.

But I agree on back plate. Thats another marketing BS. Has zero point, pretty much like those metal slabs on M.2 which are only good to keep heat there, not cool it.

I can start posting VRM close ups if there is interest, I won't promise a full breakdown.

I'm not entirely sold on the three 8 pin connectors, but I could see it being useful for extreme overclocking. I do recomend that you use at least two if you plan any significant overclock. I wish I had the kind of cooling set up neccessary to really test this.

M.2 heat sinks do more than you might think. I would agree that they aren't a requirement, and that they can't keep up with extended heavy load. Where they really excel is absorbing the heat from short bursts of heavy load and dissipate that heat over time.

 

[Mescalamba]

I can start posting VRM close ups if there is interest, I won't promise a full breakdown.

I'm not entirely sold on the three 8 pin connectors, but I could see it being useful for extreme overclocking. I do recomend that you use at least two if you plan any significant overclock. I wish I had the kind of cooling set up neccessary to really test this.

M.2 heat sinks do more than you might think. I would agree that they aren't a requirement, and that they can't keep up with extended heavy load. Where they really excel is absorbing the heat from short bursts of heavy load and dissipate that heat over time.

Even just pic of naked board is very good (there is literally no photo of board itself online).

3x 8 pin is pointless, one gives you 288W of max input power. Kinda doubt that 864W of input power would be something that this mobo could survive (also its not like 8 pins are only source of input power). Yea sure Intel has 28 core that asks for 700W (OCed a bit), but even ASUS mobo with 32-phase VRM seemed to have bit hard time (not mentioning hard time to cool that beast). Its just unrealistic, beside that there is no way to cool VRM at such power.

Im sure heatsinks are good in this case in absorbing heat, not so sure about actually dissipating it, given their pretty much flat surface. I dont get whats wrong with manufacturers lately. Heatsinks do require fins to actually work best. And its not like you cant make nice heatsink with fins..

Just checked pic of board from back side, doublers are visible so its 6x2 phase. Most likely similar to X299 XPower Gaming AC, that can give theoretical max around 720A, which in theory sounds really good. Practice is question. Since VRM would most likely overheat much sooner. If anything, 12 phase should be reasonably easy to cool. Still prefer their overkill X399 solution.

Considering its a MEG motherboard, I would expect a bit more (not in added stuff). Just in base design. X299 XPower Gaming AC is ofc pretty fine motherboard (well liked on HWbot from what I saw), just MSI had opportunity to do same as with X399 .. and they didnt do it.

 

[Black Haru]

Even just pic of naked board is very good (there is literally no photo of board itself online).

3x 8 pin is pointless, one gives you 288W of max input power. Kinda doubt that 864W of input power would be something that this mobo could survive (also its not like 8 pins are only source of input power). Yea sure Intel has 28 core that asks for 700W (OCed a bit), but even ASUS mobo with 32-phase VRM seemed to have bit hard time (not mentioning hard time to cool that beast). Its just unrealistic, beside that there is no way to cool VRM at such power.

Im sure heatsinks are good in this case in absorbing heat, not so sure about actually dissipating it, given their pretty much flat surface. I dont get whats wrong with manufacturers lately. Heatsinks do require fins to actually work best. And its not like you cant make nice heatsink with fins..

Just checked pic of board from back side, doublers are visible so its 6x2 phase. Most likely similar to X299 XPower Gaming AC, that can give theoretical max around 720A, which in theory sounds really good. Practice is question. Since VRM would most likely overheat much sooner. If anything, 12 phase should be reasonably easy to cool. Still prefer their overkill X399 solution.

Considering its a MEG motherboard, I would expect a bit more (not in added stuff). Just in base design. X299 XPower Gaming AC is ofc pretty fine motherboard (well liked on HWbot from what I saw), just MSI had opportunity to do same as with X399 .. and they didnt do it.

My thinking on the three connectors (other than marketing which is definately a factor) is to spread out the load a little. These CPUs won't pull 800 W, but they can pull over 500 W. Instead of pushing two 8 pins close to their limit, you can use three.

As for solid blocks vs fins, I'm on the side of the manufacturers for the most part. A solid block is massively cheaper to produce, and while it has only a fraction of the surface area, it still does the job.

I really wish I had the means to fairly test this. Have a VRM heat sink made to similar size spec, with 4 versions: solid copper, finned copper, solid aluminium, finned aluminium.
Run them all on the same board with the same settings and see how much difference there really is.

 

[Mescalamba]

My thinking on the three connectors (other than marketing which is definately a factor) is to spread out the load a little. These CPUs won't pull 800 W, but they can pull over 500 W. Instead of pushing two 8 pins close to their limit, you can use three.

As for solid blocks vs fins, I'm on the side of the manufacturers for the most part. A solid block is massively cheaper to produce, and while it has only a fraction of the surface area, it still does the job.

I really wish I had the means to fairly test this. Have a VRM heat sink made to similar size spec, with 4 versions: solid copper, finned copper, solid aluminium, finned aluminium.
Run them all on the same board with the same settings and see how much difference there really is.

That would be very interesting test I think.