Question Are MSI X.570 motherboard's VRMs really that bad?

[jon96789]

I have been seeing videos on YouTube about MSI's X.570 motherboards having really crappy VRMs... Other than the TOTL Godlike, Hardware Unboxed reviewed a couple of MSI boards and (MSI X.570A Pro and the MSI X.570 Gaming Edge) and their tests show that the VRM design is really bad. They show the VRM temps for both boards as being over 100-125 degree Celsius (which is even higher than the 3900X CPU), where other boards run in the 60-80 degree range. They noted that these two boards have 50% less VRM components than the Godlike which tested in the 50-60 degree range.

Because of the high temps, the 3900X had throttled back on their speed until the VRMs recovered. The VRMs reached high temps and again the CPU throttled back and this cycle was repeated.

How does this relate in real world situations (I believe they removed the heat sinks for testing on the VRMs but not sure)? I have a MSI MPG Gaming Pro Carbon and am wondering if this board suffers the same problem...

 

[DefinitelyNotTom]

What makes you think 95C is excessive?

It certainly is if you don't want the chipset fan to need to run. I finally found the temps for it... In silent mode the chipset fan comes on at 70c.

Sadly, due to these issues, I am going to spend $70 extra, most likely, and get the ace instead of pro carbon, as the ace did well in the tests. And I wasn't planning to OC, so it's mighty annoying having to spend over $300 on a mobo.

 

[jon96789]

You can a review on the MSi MPG X570 Gaming Pro Carbon Wifi here... This is the only review I found on the board. Unfortunately, they did not test the VRM capabilities...

 

[drea.drechsler]

Where is your rad mounted?
VID and Vcore are different things. You want to look at Vcore/Core Voltage.

even more precisely: Core Voltage (SVI2 TFN). Reported by HWInfo64 and a few others if you know what to look for.

VCore can be any number of things: on my motherboard it co-relates most closely to the output of the VRM, right at the capacitors and chokes (I've made measurements on my board with a digital multimeter). On some other motherboards it can be the core voltage at the LCC's at the base of the CPU. The difference can be substantial (up to 140mV) as that what makes up VDroop. That's why I say always look at SVI2 vcore.

Again, on my motherboard, SVI2 VCore co-relates very closely with measurements at the LCC's at the base of my CPU.

 

[TJ Hooker]

It certainly is if you don't want the chipset fan to need to run. I finally found the temps for it... In silent mode the chipset fan comes on at 70c.

Sadly, due to these issues, I am going to spend $70 extra, most likely, and get the ace instead of pro carbon, as the ace did well in the tests. And I wasn't planning to OC, so it's mighty annoying having to spend over $300 on a mobo.

VRM temp and chipset temp are two unrelated things. The chipset is typically located nowhere near the VRMs.

I guess there are a couple boards out there where the VRM and chipset heatsinks are connected, in which case the temp of one could affect the other. But that's rare from what I've sen.

 

[dorsai]

I have been seeing videos on YouTube about MSI's X.570 motherboards having really crappy VRMs... Other than the TOTL Godlike, Hardware Unboxed reviewed a couple of MSI boards and (MSI X.570A Pro and the MSI X.570 Gaming Edge) and their tests show that the VRM design is really bad. They show the VRM temps for both boards as being over 100-125 degree Celsius (which is even higher than the 3900X CPU), where other boards run in the 60-80 degree range. They noted that these two boards have 50% less VRM components than the Godlike which tested in the 50-60 degree range.

Because of the high temps, the 3900X had throttled back on their speed until the VRMs recovered. The VRMs reached high temps and again the CPU throttled back and this cycle was repeated.

How does this relate in real world situations (I believe they removed the heat sinks for testing on the VRMs but not sure)? I have a MSI MPG Gaming Pro Carbon and am wondering if this board suffers the same problem...

I have an MSI x570 Gaming Edge board I bought prior to the hardware unboxed video coming out...I might have made a different purchase if the video came out before I bought.

That being said I have had zero issues running a 3600 on the board with VRM temps never reaching above 60c...but on a 65w chip that's not surprising.

After the HU video I did plug in a 2700x to load the VRM and ran some all core load tests which got the VRM up to about a 100c...so the video was valid. A 100c VRM is questionable for long term use so I will be sticking with the 65w CPU's for this board.

 

[TJ Hooker]

A 100c VRM is questionable for long term use so I will be sticking with the 65w CPU's for this board.

Assuming the measured temp is close to the true hotspot temp, 100C is fine. And if one was worried, they could probably lower the temp by just increasing airflow over the VRMs.

 

[dorsai]

Assuming the measured temp is close to the true hotspot temp, 100C is fine. And if one was worried, they could probably lower the temp by just increasing airflow over the VRMs.

I agree that airflow is key...it becomes problematic though with the amount of AIOs in use these days...I stopped using one for just that reason

 

[DefinitelyNotTom]

im starting to be scared to even get the Ace now, even though it's shown to be much cooler than most. I read where a lot of people said msi is notorious for having buggy firmware updates and rush them out not tested enough. Seems very likely seeing as how they did with b450. I'd get aorus master, but it costs more since msi has rebates. And not sure if pro wifi or ultra are good enough.

 

[drea.drechsler]

.... ran some all core load tests which got the VRM up to about a 100c...

Who runs their system with such a load on it? 100C is bad, but if you never ran that 'all core load test' would it ever have experienced 100C even if you should run a 2700X? It really gets down to how you use your system: To the degree you never run your system like that it shrinks smaller and smaller as a relevant issue.

The flip side, of course can be visualized: an HPEDT work station set up in an office for everyone to do their compute-intensive tasks and therefore sees constant use in that role (like a render station, a math modeling workstation, whatever).

But to me the real issue is that these boards are just way to high-cost to put up with such low-spec VRM's.

 

[jon96789]

IMHO, at minimum I would look at the ASUS Crosshair or the Gigabyte Pro or Ultra boards. You are going to find that all the X570 boards are expensive because the chip sets are more expensive and complex which makes the engineering difficult.

 

[dorsai]

Who runs their system with such a load on it? 100C is bad, but if you never ran that 'all core load test' would it ever have experienced 100C...

The all core test in question was SETi which thousands of people run all the time so it's completely relevant as a test. Whether or not 100c is too hot for long term use remains to be seen...I'm guessing with a good number of Ryzen chips pulling way more than 65w under load we will find out the quality of those VRM designs.

 

[drea.drechsler]

The all core test in question was SETi which thousands of people run all the time so it's completely relevant as a test. Whether or not 100c is too hot for long term use remains to be seen...I'm guessing with a good number of Ryzen chips pulling way more than 65w under load we will find out the quality of those VRM designs.

My 3700X running Folding@Home only gets my B450 Mortar's VRM to 60C... even prime95 (non-AVX) got no higher than low 80's with a manual OC to 4.25Ghz. It's just crazy MSI went so far backwards in VRM design...and on expensive 570 boards at that.

And having an expensive chip isn't an excuse for cheaping out on the VRM. In fact, it makes it even more important to provide value everywhere on the board to justify the extra cost.

 

[jon96789]

Sadly, here is another online review which blasts MSi's crappy VRM design... KitGuru tested four $200 boards and the MSi really stunk... The MSi board ran a whopping 40 degrees hotter than the others on the HWInfo64 app and 30-40 degrees hotter with the thermo-coupled measurements.. They ran the test pushing the CPU to 4.25 GHz @ 1.35 volts. The YouTube video can be found here.

Compared with other boards in the review...

They end with the quote:
MSI’s MPG X570 Gaming Edge WiFi uses basic MOSFETs and tries to use brute force to cool them – but this doesn’t work very well, with sensor-reported temperatures as high as 110°C and physical PCB readings showing more than 100°C. I wouldn’t recommend this board for use with an overclocked Ryzen 12-core or 16-core CPU. That said, it’s likely to be fine for lower-powered parts, which is a good thing as I have previously recommended this exact model for a friend’s Ryzen 7 3700X build.

 

[drea.drechsler]

They end with the quote:
.... That said, it’s likely to be fine for lower-powered parts, which is a good thing as I have previously recommended this exact model for a friend’s Ryzen 7 3700X build.

the bottom line: it's good enough even for 8 core so it's hardly "hot garbage" as someone said.

But $200 is a price point reserved for boards you can say a lot more about than "at least it ain't trash"

 

[jon96789]

Well, Hardware Unboxed released their final X570 Motherboard review, on boards priced at $250... Crap! The board I have, the MSi MPG X570 Gaming Carbon Pro WiFi turned out to be the worst of all the X570 boards tested so far. The board thermal throttled the CPU when the VRM temps hit 95 degrees C.

On the good news, they reported that MSi is releasing a new board, the MSi MEG X570 Unify at the end of October. This $300 board has the same VRM design as the MSi MEG X570 Ace board with better heat sinks. On their testing, the board VRMs only reached around 65 degrees which is excellent, beating the Ace by a couiple of degrees.

The bad news is that I now have to buy a new board to replace the piece of junk that I have. I have a AMD Ryzen 9 3900X and the VRMs are always hitting 95 degrees under load which means it is throttling the CPU at that point (the clock speed at 95 degrees drops from 4.2 GHz to 3.7 GHz, not much but is still a drop)

 

[drea.drechsler]

.......The board thermal throttled the CPU when the VRM temps hit 95 degrees C.....

That sounds like a darn good idea to me.... I'd much rather it throttle than start smoking.

Just saying that sort of thing in isolation is pretty much pointless: things an inquiring sort would be asking: what CPU, under what load for how long. Also, what sort of cooling on the CPU and what sort of airflow around the VRM area since free-air test benches with liquid cooling on the CPU provide terrible airflow across the VRM if not provided for.

EDIT add: watched most of the video... it sounds to me like you should remove the plastic bits and make sure your case provides airflow across the VRM area. Steve makes a point he tested in open-air not in a proper case.

But, as always, I think it's shameful to be selling boards at this price point with such weak VRM's... especially since it appears all it needs is a better heatsink to make it first rate. No major design effort that, ferchrisake.

 

[jon96789]

I saw a major flaw for the MSi Unify... It appears to be a stripped down MSi Ace board. As such, it has the same four SATA port limitation...

 

[jon96789]

I decided to run a different stress test on my system. Since I bought the PC for video encoding, I decided to encode a video in the highest quality (SHQ) setting in Handbrake.

This encode takes one hour with the AMD 3900X (it used to take 150 minutes with my old Intel i7 6700K CPU. Right now, the encode has been running for 30 minutes. The stats are as follows:

CPU speed maximum:	4.3 GHz
CPU Speed average:	4.05 GHz
CPU temp maximum:	89.9 degrees
CPU temp average:	79.8 degrees
VR Temp maximum:	84.0 degrees
VR Temp average:	80.2 degrees
X570 temp maximum:	73.3 degrees
X570 temp average:	72.6 degrees

Since this is the most stressful task i will be doing, I decided to use this as a baseline to see if i really need to replace the motherboard. I guess the max temp of 83 degrees for the VR while very high, it is still within thermal limits, especially when the MSi Ace or Unify boards run only 65 degrees under heavier stress loads.

 

[jon96789]

I have been encoding a video for about two hours with Handbrake... HWiNFO64 indicates that my CPU has thermal throttled. The VR temps during the encoding never surpassed 84 degrees C. The CPU speed during that time peaked at 4.275 GHz but dropped down to 3.32 GHz during the encode.

I launched HWiNFO64 about 30 minutes into the encoding so the CPU should have been running at full speed since then.

 

[jon96789]

Talking with MSi tech support got me nowhere. They are saying if the CPU is throttling, the CPU is defective. If they were to RMA my board, they would just send me an identical refurbed board. That is not going to solve my issue of the CPU thermal throttling whenever I am encoding videos.

Today, I decided to bite the bullet and buy the ASUS ROG Crosshair VIII Hero (sans WiFi) board from Best Buy for $360. They already shipped it and I should get it on Friday. The ASUS board was the best sub $400 board according to the Hardware Unboxed videos. The top three VRM performing boards were all $700-$1000. The ASUS was almost as good as the top three boards, only a few degrees warmer than the other three.

Still a 60 degree difference from the MSi MPG boards is pretty indicative of how bad the MSi VRM design is. The MSi boards may be fine with the AMD 65-watt CPU but it definitely cannot handle the demands of the AMD 105-watt CPUs, such as the 3800x/3900x and especially the 3950x which has four more cores than the 3900x.

I only hope than I can successfully transplant the motherboard without reinstalling everything from scratch.

 

[TJ Hooker]

HWiNFO64 indicates that my CPU has thermal throttled

From your own description it's your CPU temps that are the problem, not VRM temps. Why replace the motherboard?

 

[jon96789]

After I updated to the latest MSi BIOS, the profile seems to have changed. The CPU averages 80-85 degrees and the VRMs varies from 85-95 degrees. The latest version of HWiNFO64 now indicating the system is thermal throttling when under load. Hardware Unboxed has stated that this particular board throttles the CPU at relatively low temps compared to other MSi boards.

 

[TJ Hooker]

It is normal for the CPU to start incrementally dialing back the boost clock starting somewhere around 70-80C (can vary by BIOS version). Maybe hwinfo reports that as thermal throttling?

https://www.tomshardware.com/news/amd-ryzen-3000-boost-frequency-bios-fix-agesa,40359.html

 

[jon96789]

I just got my ASUS X570 Crosshair VIII Hero board today and swapped out the MSi X570 MPG Gaming Carbon Pro WiFi... Ran into some unexpected problems during the install. First, one of my RAM modules went bad before the install and caused a bunch of blue screens. Ran a memory test and it returned a bad RAM module error so I had to remove two 8 GB modules and am now down to 16 GB until I get the RAM RMA'd from Corsair.

Before I did the swap, I booted into Windows and uninstalled all the unnecessary MSi drivers. Then right after I did the reboot, Microsoft started installing Windows 10 v.1909, which took a extremely long time, over one hour.

I swapped out the motherboard, installed the M.2 NVMe SSD and the optical and other hard drives. Reconnected my Corsair H115 RGB Platinum cooler and booted up... I now keep getting a CPU fan error, the BIOS does not see any CPU fan. The Corsair fans are connected to the USB hub, not the CPU fan connectors on the motherboard. I have to figure that one out.

I am now running Prime 95 with HWiNFO64... HWiNFO64 does NOT recognize the VRM sensors on the motherboard so now I have no way of determining how hot the VRMs are getting. Using an infrared thermal thermometer, it shows the VRMs at 50 degrees C, a LOT cooler than the MSi.

The CPU temp averages about 65-70 degrees, cooler than before, it peaked at 95 degrees for a few seconds. The big difference was the CPU speed. The board has the original BIOS dated 0507. The CPU with that BIOS peaked on a couple of cores at 4.57 GHz, close to the rated CPU speed, which is about 200 MHz higher than the MSI board. The rest of the cores hit 4.3 GHz, but it settles down to 4.00 GHz, the MSi board settles down to 3.9 Ghz.

 

[jon96789]

My grandson needed a desktop PC, since I had the MSi X570 MPG Gaming Pro Carbon WiFi sitting around along with an unused case and 16GB RAM, I decided to build a PC. All I had to get was a GPU and a CPU and a hard drive.

I went with the AMD Ryzen 7 3700X and a NVidia GTX1650 along with a Samsung 960 SSD. I put everything together and this is what I found out running Prime95. The motherboard works fine with a 65-watt CPU. The 3700X peaked at 4.35 GHz, a tad short of the rated 4.4 GHz. The CPU temps averaged about 65-70 degrees with the stock AMD Wraith Prism air cooler. More importantly, the VRMs barely blipped, reading 45 degrees C, a heck of drop from the 3900X's 90-95 degrees with this board.

Astonishingly, the new build boots in only 30 seconds whereas my current ASUS X570 ROG Crosshair VIII Hero takes about 65-70 seconds. The BIOS boot on the MSi is suddenly shorter and boot time on the SSD is far faster than my Samsung M.2 NVMe SSD.

In AMD Balanced power mode the CPU idles at ~50 degrees and in Power Saver mode it idles at 35 degrees.