MSI's M.2 "Heat Shield" Increases SSD Temperature

[TheNeonWhiteOne]

According to gamer nexus MSI's heat shield for m.2 ssds actually increase temperatures of the ssd.  The shield ends up trapping the heat causing the ssd to be hotter. 

http://www.gamersnexus.net/guides/2781-msi-m2-heat-shield-increases-temperwtures

[BingoFishy]

You need a quote and personal opinion.

[Wolther]

I watched this earlier today. Yeah it makes sense if you think about it. There no fins, has a pretty terrible heat pad. Also it doesn't even surround the entire m.2 (which causes heat to be trapped in certain pockets). Resulting a possible increase in the ssd temperature 

 

Maybe I missed it, but I wish he did testing on if it reduced the heat when taking into account a gpu being above it. Perhaps it'll protect the ssd somewhat from the air being blown off of the GPU? 

[Ensho]

Yep.  Watched that video and it was very informative.  Steve does a good job of actually reviewing stuff and not just being an extended branch of companies marketing departments.

[mattebad]

Honestly this was not GN best video, the whole point of the heat shield is to shield from heat and act as an insulant. The point of the heat shield is to prevent GPU fans from venting out heat onto the SSD and potentially damaging it over long term. Why GN didn't test with a gpu which is what the whole purpose of the heat shield is for, idk, maybe it was just an oversight. Just thought to add some context here

[AHaskin14]

So there's no direct contact to the SSD....... Just an air gap...... That's smart

[cj09beira]

35 minutes ago, Wolther said:

I watched this earlier today. Yeah it makes sense if you think about it. There no fins, has a pretty terrible heat pad. Also it doesn't even surround the entire m.2 (which causes heat to be trapped in certain pockets). Resulting a possible increase in the ssd temperature 

 

Maybe I missed it, but I wish he did testing on if it reduced the heat when taking into account a gpu being above it. Perhaps it'll protect the ssd somewhat from the air being blown off of the GPU? 

the air from an gpu helps, the air it self when exiting the gpu is not hot, at least not 45c-65c hot, so that air helps.

and the main problem of the shield is how it blocks the sides, and that is the problem, it prevents air from excaping the bottom side.

[Notional]

It's so dumb. PCPer already proved that it's almost practically impossible to throttle an NVME SSD with no active cooling, and that it is completely impossible to throttle it when there is active cooling present. Even when that cooling, is hot exhaust from a GPU directly on it.

This is a gimmick, but a gimmick that does more harm than good. Worst part is, that the customers of this board will probably use it.

[Wolther]

8 minutes ago, cj09beira said:

the air from an gpu helps, the air it self when exiting the gpu is not hot, at least not 45c-65c hot, so that air helps.

and the main problem of the shield is how it blocks the sides, and that is the problem, it prevents air from excaping the bottom side.

Air that's hotter than a component does not help that component cool off. You m.2 isn't gonna be under full load 24/7, your gpu is more likely to be under load than the m.2. 

Like I said, testing on this would be nice, rather than just taking your assumption on how much it affects a SSD. 

[cj09beira]

1 minute ago, Wolther said:

Air that's hotter than a component does not help that component cool off. You SSD isn't gonna be under full load 24/7, your gpu is more likely to be under load than the SSD. 

Like I said, testing on this would be nice, rather than just taking your assumption on how much it affects a SSD. 

it would only make a difference when the ssd is hot, when its not it doesn't matter if its 3-4 degrees higher.

but i agree we need a test.

[dizmo]

While it's unfortunate, I'm not too surprised. I'd still use it, because I always saw it as more of an aesthetic improvement than  performance one.

 

11 minutes ago, Notional said:

 

This is a gimmick, but a gimmick that does more harm than good. Worst part is, that the customers of this board will probably use it.

Gimmick for cooling, maybe, but still an improvement for some in the aesthetics department.

[SpaceGhostC2C]

Rather besides the topic, but I was a bit surprised by the statment

 

Quote

That’s an improvement of 1C for the top of the M.2 SSD, and because of how we’re testing, that’s not “within margin of error,” as folks seem to like to spout without understanding. That is repeatable and provable, and our equipment has been calibrated in thermal chambers.

 

It's perfectly possible that their instruments and methodology can distinguish 1C differences. However, the way to prove that 1C is not within margin of error is, you know, provide your margin of error, as small as it may be. Not by making random comments about calibration and what others "sprout without understanding". Unless they don't really know themselves what degree of precision they actually have, or are trying to claim that their margin of error is exactly zero, in which case they are simply bullshitting (or "sprouting without understanding").

 

Back on the main point, I would have expected the idea of the same object being a good conductor (for dissipation) and a bad conductor (for isolation) to find some resistance within the company before even hitting the market...

[Notional]

21 minutes ago, Wolther said:

Air that's hotter than a component does not help that component cool off. You m.2 isn't gonna be under full load 24/7, your gpu is more likely to be under load than the m.2. 

Like I said, testing on this would be nice, rather than just taking your assumption on how much it affects a SSD. 

 

The Samsung 950 Pro did not thermal throttle with hot GPU exhausting on it. It only throttles, when it reaches 90c or so. If your GPU exhaust is that hot, you're doing it wrong.

 

13 minutes ago, dizmo said:

Gimmick for cooling, maybe, but still an improvement for some in the aesthetics department.

 

Aesthetics that can cause your very expensive SSD to throttle. What's the point then? Don't buy ugly green PCB SSD's

[Wolther]

1 minute ago, Notional said:

The Samsung 950 Pro did not thermal throttle with hot GPU exhausting on it. It only throttles, when it reaches 90c or so. If your GPU exhaust is that hot, you're doing it wrong.

 

 

Haven't said anything about thermal throttling. Temperature can affect the longevity of drives, also people just don't like seeing their components get hot. We could have a long discussion on whether it matters or not, but that's just how it is. Some people are fine with their GPUs are 80 C constantly, some people say 70 C is max for them. It's just whatever someone wants for their system. 

[dizmo]

2 minutes ago, Notional said:

 

Aesthetics that can cause your very expensive SSD to throttle. What's the point then? Don't buy ugly green PCB SSD's

I'd just make sure I didn't get an m.2 drive that ran hot to begin with
IIRC the Hyper X they used isn't that great. It chewed through the battery test Tom's does compared to higher end Samsung stuff.

[Notional]

1 minute ago, Wolther said:

Haven't said anything about thermal throttling. Temperature can affect the longevity of drives, also people just don't like seeing their components get hot. We could have a long discussion on whether it matters or not, but that's just how it is. Some people are fine with their GPUs are 80 C constantly, some people say 70 C is max for them. It's just whatever someone wants for their system. 

 

Core temp, and exhaust temps are not the same. The latter is much much lower. The operating temp of a 950 Pro is 1-70c with thermo throttle around 90c (so much for 70c). If your GPU gets to 80-90c the exhaust temps will be under 70c. It's not going to have any effect on the SSD. It will be obsolete and have dead NAND before temps kill it off anyways.

In the end this thing does more damage then good. And you pay extra for the privilege.

[Notional]

3 minutes ago, dizmo said:

I'd just make sure I didn't get an m.2 drive that ran hot to begin with
IIRC the Hyper X they used isn't that great. It chewed through the battery test Tom's does compared to higher end Samsung stuff.

 

Then you don't want a fast NVME. The controller chip gets very hot on those. It's why they need a little active cooling, when they are being hammered to their max.

[Wolther]

3 minutes ago, Notional said:

Core temp, and exhaust temps are not the same. The latter is much much lower. The operating temp of a 950 Pro is 1-70c with thermo throttle around 90c (so much for 70c). If your GPU gets to 80-90c the exhaust temps will be under 70c. It's not going to have any effect on the SSD. It will be obsolete and have dead NAND before temps kill it off anyways.

In the end this thing does more damage then good. And you pay extra for the privilege.

Never said core temp and exhaust temp is the same thing. It's obviously not since exhaust temp is spread out in a bigger volume. 

Your information your I guess trying to influence me with doesn't tell me anything, seeing as how your making assumptions and not doing real-world testing. I would like to see GN do testing when considering that "shielding" part of this component. As how I think that was the whole point of the product but completely skipped over it. 

[Doobeedoo]

I've read about this today, and that the test was done via benchmark stress over 60min hmmm. Odd. Though I get it's not shielded from both sides and that it's rather thin shield too. Possibly good to counter GPU heat. But as said using stressing benchmark over long periods of time which would write a whole drive numerous times in max speed makes no sense, specially for consumer user.

[dizmo]

16 minutes ago, Notional said:

Then you don't want a fast NVME. The controller chip gets very hot on those. It's why they need a little active cooling, when they are being hammered to their max.

The 960 EVO has added heatsink features. So under load it hits 76, thus under the shield it'd hit 80. Still well below the 93 this one was hitting. As with all high end builds it's all about proper component selection.

His test definitely shows how it increases performance, but it doesn't mean it's wrong to use it in all scenarios.

[Notional]

12 minutes ago, Wolther said:

Never said core temp and exhaust temp is the same thing. It's obviously not since exhaust temp is spread out in a bigger volume. 

Your information your I guess trying to influence me with doesn't tell me anything, seeing as how your making assumptions and not doing real-world testing. I would like to see GN do testing when considering that "shielding" part of this component. As how I think that was the whole point of the product but completely skipped over it. 

 

https://www.pcper.com/reviews/Storage/Samsung-950-PRO-256GB-and-512GB-M2-NVMe-PCIe-SSD-Review/Thermal-Throttling-Conclusio

 

On the podcast Allyn elaborated, that even exhaust from a GPU would prevent thermal throttling completely. This shield has no purpose, as heat doesn't harm these SSD's at the exhaust temps you would see inside a pc. However this shield will act as an insulator, causing the SSD to heat itself up to throttling temps, it wouldn't otherwise reach.

 

5 minutes ago, dizmo said:

So under load it hits 76, thus under the shield it'd hit 80

That's an assumption you have no empirics to back up.

[dizmo]

37 minutes ago, Notional said:

That's an assumption you have no empirics to back up.

I'd rather go with educated guess
That is how hot the 960 gets under load. If the shroud adds 4 degrees to the Hyper X, I doubt it'd add more than 4 degrees to drive with better thermals. Even so, if it adds 10 degrees that's still well under the throttling point.

 

Also you didn't use that word properly.

 

[zMeul]

it's called a heat shield for a reason 

 

and here's one thing the Gamers Nexus article and video misses - where are the IO Meter numbers for heat shield off and on?

did the heat shield actually helped with worse performance? or did the temp numbers got higher because the SSD showed better IO numbers?

[Energycore]

3 hours ago, TheNeonWhiteOne said:

According to gamer nexus MSI's heat shield for m.2 ssds actually increase temperatures of the ssd.  The shield ends up trapping the heat causing the ssd to be hotter. 

http://www.gamersnexus.net/guides/2781-msi-m2-heat-shield-increases-temperwtures

Hi there! Please format your post so that it complies with the Tech News and Reviews guidelines. Include the most relevant quote from the article and preferably also the link to the youtube video that GN did.

 

Also give us your opinion to spark the discussion.

[Vode]

On 1/27/2017 at 11:47 PM, mattebad said:

Honestly this was not GN best video, the whole point of the heat shield is to shield from heat and act as an insulant. The point of the heat shield is to prevent GPU fans from venting out heat onto the SSD and potentially damaging it over long term. Why GN didn't test with a gpu which is what the whole purpose of the heat shield is for, idk, maybe it was just an oversight. Just thought to add some context here

It's hard to call it a heat shield when it's metal and a thermal pad.

 

If the shield is hotter than the SSD itself then it will transfer the heat. 

 

The term Heat shield is just wrong.