Thursday, February 2nd 2017

MSI M.2 Shield is Snake Oil Say Tests, Company Refutes Charges

With its 200-series chipset motherboards, MSI introduced its exclusive M.2 Shield accessory on some of its more premium products. This aluminium heatspreader with a thermal pad, according to its makers, is designed to lower temperatures on M.2 SSDs, and reduce thermal throttling of performance. Tests by Gamers Nexus claim that far from reducing throttling, the M.2 Shield creates a "heat trap" that throttles performance further.

The M.2 Shield accessory was tested by Gamers Nexus using a Kingston HyperX Predator PCIe M.2 drive, on an MSI Z270 Gaming Pro Carbon motherboard. Driven by a Marvell 88SS9293 controller, the drive is known to heat up and lose performance to overheating. The data presented by the publication is Delta T (temperatures subtracting ambient temperature). MSI on its part, stands by the efficacy of the M.2 Shield accessory, and is sharing testing methods and data of its own with media sites. Tests by other publications such as Bit-Tech and KitGuru show positive results for the M.2 Shield.
Sources: GamersNexus, ComputerBase.de
Add your own comment

36 Comments on MSI M.2 Shield is Snake Oil Say Tests, Company Refutes Charges

#1
Dimi
I have no idea what this title means. Anyone care to explain?
Posted on Reply
#3
9700 Pro
Shield is a good name for it, since it prevents the heat getting out. :toast:

Just like those Thermal Armors on Asus' TUF series, though I was too lazy to rip that off from my old Sabertooth Z87..
Posted on Reply
#4
pat-roner
Well, as Steve from Gamers Nexus mentioned, the reason it has a bad effect on the drive is not because the top side isn't reduced, it's that it traps hot air under the drive, and for drives with chips on both sides it has a negative effect. Both drives that Bit-Tech and Kitguru used are "single sided" drives, and thus there are no chips heating up the underside.
Posted on Reply
#5
TheGuruStud
I guess a real heatsink with fins was just too damn complicated (that doesn't block airflow).
Posted on Reply
#6
natr0n
Heatsinks need airflow to work.

That fool seems more into his hair than thinking.
Posted on Reply
#7
TheGuruStud
natr0n said:
Heatsinks need airflow to work.

That fool seems more into his hair than thinking.
I hate to break it to you, but that's not true (you mean by fan). This isn't a heatsink anyway, so how can it work lol.
Posted on Reply
#8
VSG
pat-roner said:
Well, as Steve from Gamers Nexus mentioned, the reason it has a bad effect on the drive is not because the top side isn't reduced, it's that it traps hot air under the drive, and for drives with chips on both sides it has a negative effect. Both drives that Bit-Tech and Kitguru used are "single sided" drives, and thus there are no chips heating up the underside.
Kitguru using thermal imaging was enough to make me close that tab. There is really no point using this to quantify heat at this level without emissivity accounted for, and a high precision probe used. This was also the case when the EVGA GTX 1080 FTW VRM issues popped up, with thermal imaging results being used to say they ran too hot when actual thermocouple measurements proved otherwise.
Posted on Reply
#9
pat-roner
natr0n said:
Heatsinks need airflow to work.

That fool seems more into his hair than thinking.
You don't think theres airflow in a case? No case har 100% still air.
Posted on Reply
#10
natr0n
This guy used a test bench and even admits you need a side case fan. He didn't even simulate airflow over it. Flawed review.

Now... lets argue about it.
Posted on Reply
#11
Hood
MSI has always been substandard, why is this surprising? I guess they do serve a purpose, though - teaching newbies how to pick hardware - after being burned with an MSI board, maybe they'll do better next time and buy Asus or Asrock...
Posted on Reply
#12
Steevo
VSG said:
Kitguru using thermal imaging was enough to make me close that tab. There is really no point using this to quantify heat at this level without emissivity accounted for, and a high precision probe used. This was also the case when the EVGA GTX 1080 FTW VRM issues popped up, with thermal imaging results being used to say they ran too hot when actual thermocouple measurements proved otherwise.
I be the people that had their VRM explode and catch fire would disagree. Youtube: WAbl0fLY06U And the fact the manufacturer provided thermal pads and an updated BIOS..... it must mean nothing.
Posted on Reply
#13
D1RTYD1Z619
Hood said:
MSI has always been substandard, why is this surprising? I guess they do serve a purpose, though - teaching newbies how to pick hardware - after being burned with an MSI board, maybe they'll do better next time and buy Asus or Asrock...
I agree I have never considered MSI a top tier company.
Posted on Reply
#14
VSG
Steevo said:
I be the people that had their VRM explode and catch fire would disagree. Youtube: WAbl0fLY06U And the fact the manufacturer provided thermal pads and an updated BIOS..... it must mean nothing.
Did you read the whole thing about the VRMs? It was faulty modules, not the cooler, that was responsible. The thermal pad kit was purely for PR. It does help reduce temps, but the stock solution was plenty enough for the modules in the first place.
Posted on Reply
#15
uuuaaaaaa
VSG said:
Did you read the whole thing about the VRMs? It was faulty modules, not the cooler, that was responsible. The thermal pad kit was purely for PR. It does help reduce temps, but the stock solution was plenty enough for the modules in the first place.
IIRC Buildzoid actually said that it was probably a bad batch of caps. Gamers Nexus and Buildzoid did a colab on that topic.
Posted on Reply
#16
Steevo
VSG said:
Did you read the whole thing about the VRMs? It was faulty modules, not the cooler, that was responsible. The thermal pad kit was purely for PR. It does help reduce temps, but the stock solution was plenty enough for the modules in the first place.
"Modules" Like... Voltage Regulation Modules..... http://forums.evga.com/1080-FTW-exploded-now-dead-and-took-other-parts-with-it-m2583259.aspx https://www.extremetech.com/computing/238633-evga-gtx-1080s-1070s-allegedly-exploding-due-improper-vrm-cooling

Congrats on obviously reading it yourself.
Posted on Reply
#17
VSG
Steevo said:
"Modules" Like... Voltage Regulation Modules..... http://forums.evga.com/1080-FTW-exploded-now-dead-and-took-other-parts-with-it-m2583259.aspx https://www.extremetech.com/computing/238633-evga-gtx-1080s-1070s-allegedly-exploding-due-improper-vrm-cooling

Congrats on obviously reading it yourself.
I did, yes. Bad batch of modules (VRMs and chokes alike) which affected some cards. Unfortunate for sure, but it wasn't the cooler at fault. Anyway this is off topic to the conversation here.
Posted on Reply
#18
Chaitanya
If anyome wants a good M.2 ssd cooler just dont be a pennypincher get a Aquacomputer kryo m.2 or Angelbird wings px2. Msi and asus(maximus code, formula, tuf, etc) both with ssd shield are just plain garbage solutions.
Posted on Reply
#19
Caring1
Dimi said:
I have no idea what this title means. Anyone care to explain?
Snake oil is the term used to describe any suspect or bogus remedy.
Posted on Reply
#20
Steevo
VSG said:
I did, yes. Bad batch of modules (VRMs and chokes alike) which affected some cards. Unfortunate for sure, but it wasn't the cooler at fault. Anyway this is off topic to the conversation here.
It is off topic to a degree. But you are wrong, the article clearly states that the 200+F degree VRM temps were not an "issue" but they will provide thermal pads, a new BIOS and if you felt uncomfortable installing those items a card would be cross-shipped to you. If this were a car for example, and it ran fine in the driveway, in town, but as soon as you got on the interstate it overheated some, and may possibly burst into flames and burn up, but they will provide a new radiator and change the fan speed with software.... but nothing is wrong, just don't drive it on the interstate until you get this, "not fix" for the "not problem".

You see the defect in the design yet? No chokes were exploding, they are just copper wrapped around a ferrite core, nothing to explode or burn, the VRM's were dying either due to not being able to handle the temperatures hot enough to boil water, which means they used VRM's with specs below what was engineered (common in manufacturing) or the solder was melting and shorting out (again cheaper solder used in production could be the cause) but the eventual fix for the issue was a higher fan speed and pads to transfer the heat away through the backplate and some contact with the cooler fins.
Posted on Reply
#21
VSG
Steevo said:
It is off topic to a degree. But you are wrong, the article clearly states that the 200+F degree VRM temps were not an "issue" but they will provide thermal pads, a new BIOS and if you felt uncomfortable installing those items a card would be cross-shipped to you. If this were a car for example, and it ran fine in the driveway, in town, but as soon as you got on the interstate it overheated some, and may possibly burst into flames and burn up, but they will provide a new radiator and change the fan speed with software.... but nothing is wrong, just don't drive it on the interstate until you get this, "not fix" for the "not problem".

You see the defect in the design yet? No chokes were exploding, they are just copper wrapped around a ferrite core, nothing to explode or burn, the VRM's were dying either due to not being able to handle the temperatures hot enough to boil water, which means they used VRM's with specs below what was engineered (common in manufacturing) or the solder was melting and shorting out (again cheaper solder used in production could be the cause) but the eventual fix for the issue was a higher fan speed and pads to transfer the heat away through the backplate and some contact with the cooler fins.
My point when I brought this up was that I do not trust thermal imaging as a means of quantifying heat, and so I am taking the Toms Hardware numbers with a grain of salt. The exact VRM modules used on those cards are rated at well above 125 °C, including the ones that went bad. Gamers Nexus used multiple thermocouples with EMI shieldingand showed that temps were well within rated specs with and without the new BIOS and thermal pad kit. To bring this back to topic, I argue the same applies to the M.2 heatsink test here- thermal imaging done by Kitguru is not something I agree with, so I am more reliant to take the Gamers Nexus tests as a point of reference.

That said, I did note some things that need to be re-done and I have been having discussions with Steve about this since Monday. So that Gamers Nexus article will be updated shortly. Similarly, I can definitely understand the skepticism behind the EVGA VRM issue and I wholeheartedly agree that the issue was not resolved to my satisfaction either. My contacts at EVGA mentioned it may have been bad modules, but without the actual data it is left to their word.
Posted on Reply
#22
peche
Thermaltake fanboy
well, the cover might no be a problem if:
Airflow could blow it and help to reduce temps, cuz magically it won't cool down,
some fins will be better than a flat surface, for heat exchange, no matter if the fins are 1mm or 10 cm, the matter it's like physically heat tends to raise, so will be easier to cool down
Also location of the drive with the shield will be affected if some elements will block airflow,
and maybe the most important part, thermal pads, not all M.2 drives have the same size, so thermalpads may not fit all drives if they are slim or thicker compared to the supplied thermalpad size,
i prefer to add manually add or stick regular Heatsinks, or ram heatsinks to a single M.2 drive instead that shield....
Posted on Reply
#23
Steevo
VSG said:
My point when I brought this up was that I do not trust thermal imaging as a means of quantifying heat, and so I am taking the Toms Hardware numbers with a grain of salt. The exact VRM modules used on those cards are rated at well above 125 °C, including the ones that went bad. Gamers Nexus used multiple thermocouples with EMI shieldingand showed that temps were well within rated specs with and without the new BIOS and thermal pad kit. To bring this back to topic, I argue the same applies to the M.2 heatsink test here- thermal imaging done by Kitguru is not something I agree with, so I am more reliant to take the Gamers Nexus tests as a point of reference.

That said, I did note some things that need to be re-done and I have been having discussions with Steve about this since Monday. So that Gamers Nexus article will be updated shortly. Similarly, I can definitely understand the skepticism behind the EVGA VRM issue and I wholeheartedly agree that the issue was not resolved to my satisfaction either. My contacts at EVGA mentioned it may have been bad modules, but without the actual data it is left to their word.
Photon emission of a specific wavelength is exactly what the cameras are tuned for, filtered for, and if you want, the background is listed at 72F and emission of more photons is a function of temperature. Its not magic, and most thermal cameras are accurate to within a degree or two, which is as good as you can get with a thermocoupler that is a "dumb" and uncalibrated device that suffers from manufacturing defects as much as anything else. A thermal camera will tell you about a localized hot spot much better than any form of probe, which by default acts as an auxiliary heat-sink.

http://www.flir.com/science/blog/details/?ID=74935

"As we have seen, the RSS uncertainty analysis technique allows us to determine the accuracy of infrared cameras, and that these cameras may have, at most, a 2ºC margin of error. With proper calibration and attention to factors such as ambient temperature, emissivity, and spot size, the possible margin of error can be less than 1ºC."

Source sauce.

https://www.picotech.com/library/application-note/improving-the-accuracy-of-temperature-measurements#thermistors

"Thermocouples are not, however, precision sensors: errors of 2 °C are typical."
Posted on Reply
#24
VSG
Steevo said:
Photon emission of a specific wavelength is exactly what the cameras are tuned for, filtered for, and if you want, the background is listed at 72F and emission of more photons is a function of temperature. Its not magic, and most thermal cameras are accurate to within a degree or two, which is as good as you can get with a thermocoupler that is a "dumb" and uncalibrated device that suffers from manufacturing defects as much as anything else. A thermal camera will tell you about a localized hot spot much better than any form of probe, which by default acts as an auxiliary heat-sink.

http://www.flir.com/science/blog/details/?ID=74935

"As we have seen, the RSS uncertainty analysis technique allows us to determine the accuracy of infrared cameras, and that these cameras may have, at most, a 2ºC margin of error. With proper calibration and attention to factors such as ambient temperature, emissivity, and spot size, the possible margin of error can be less than 1ºC."

Source sauce.

https://www.picotech.com/library/application-note/improving-the-accuracy-of-temperature-measurements#thermistors

"Thermocouples are not, however, precision sensors: errors of 2 °C are typical."
That accuracy for the thermal camera is when you know exactly what the emissivity is for the object you are measuring. Are you telling me that Kitguru, Toms etc all did this measurement for the exact SSD/VRM/PCB each time?

Also, the thermocouples I had recommended Steve get have a rated accuracy of < 1 °C which I felt was plenty enough for his tests. Again, I would love to be corrected about this as it will save me a lot of trouble in my research lab, but there's a reason all of us at my university use thermocouples and thermistors for our heat transfer measurements and that is consistent accuracy and high precision for our purposes- something a consumer level thermal camera has not provided yet to my knowledge.

Edit: Just got done with a discussion with some thermal engineers/friends at Corsair, Asetek and Intel and they all were saying the same thing. I will contact FLIR also because I am genuinely curious and want to make sure I am not overlooking something :)

Double edit: Just got off the phone with the local FLIR tech rep, and basically he said the same thing. The consumer level gadgets from their gadgets division are a no-go for quantification, but higher end test bench kits with Kapton tape on the component of interest followed by calibdation can be accurate to within 3-4 °C. Not bad at all, but the minimum he recommended was a $6k USD item. He isn't a sales rep so I don't believe that had anything to do with the recommendation of the expensive kit. Toms Hardware seems to have a really good IR setup, and I imagine they must also have the expertise/knowledge to use said tape or similar functioning alternative too. As such, I feel more confident about the Toms Hardware measurements now which applies to the EVGA VRMs (again their measurements from their thermal camera were within spec of the VRMs), but no info on Kitguru's (Luke Hill) measurements yet. A different Kitguru reviewer uses an unnamed Fluke IR camera which doesn't help much either. Will keep looking!
Posted on Reply
#25
Steevo
I'm not trying to be an asshole, I work with precision GPS for a living, and understand the difference in quality. Sometimes trying to explain why an oscilloscope is needed to diagnose a problem to an average tech spills over here.
Posted on Reply
Add your own comment