PDA

View Full Version : Can M.2 SSD's really get this hot or is it fake?



FalloutBoy
09-28-2015, 10:39 AM
Question as above, is this click bait or true?

https://www.youtube.com/watch?v=3hhdWwvh5kI

joshindaphils
09-28-2015, 11:02 AM
Lets apply Occam's razor here.

A reputable long time custom system builder has purchased a very expensive FLIR camera and went to considerable effort to produce a fraudulent video as click bait (on an obscure topic none the less) netting less that nine thousand views in over a year.

OR

It is true.

FalloutBoy
09-28-2015, 11:38 AM
Lets apply Occam's razor here.

A reputable long time custom system builder has purchased a very expensive FLIR camera and went to considerable effort to produce a fraudulent video as click bait (on an obscure topic none the less) netting less that nine thousand views in over a year.

OR

It is true.

Personally I though it was true also but was told by the chap who was producing the 3D printed drive cages here that there was no way it would
ever get to this temp via a PM. I just wanted to know one way or the other as I was planning on putting the M.2 overlapping a drive with a max
of 70.c so if it reaches 106.c it's not going to be very good for it's health and / or any plastic printed part one would think unless it's thermoplastic
with a very high temp tolerance.

Dr. Zchivago
09-28-2015, 12:17 PM
Lets apply Occam's razor here.

A reputable long time custom system builder has purchased a very expensive FLIR camera and went to considerable effort to produce a fraudulent video as click bait (on an obscure topic none the less) netting less that nine thousand views in over a year.

OR

It is true.

There are other possibilities:
1) Assuming the same system for both drives, there exists the possibility of a faulty temp sensor.
2) The monitoring software could be in conflict with another process (as when a system with multiple monitoring softwares sees CPU temps in the negatives while air cooling in 25 C ambient).
3) Some initial parameter not discussed by the builder was ignored for one build or the other - I didn't see an excerpt for methodology or photographs of the testing setup. Even pros make mistakes.
4) Something I didn't think of in the last three minutes.

Point: There are always more than two possibilities. Knowing what I know (admittedly, less than a hardware engineer and more than the average user), I would also be suspicious of readings that high on hardware that the manufacturer's data sheet lists as having an operational temp range between 0 C and 70 C.

Z

joshindaphils
09-28-2015, 12:48 PM
There are other possibilities:
1) Assuming the same system for both drives, there exists the possibility of a faulty temp sensor.
2) The monitoring software could be in conflict with another process (as when a system with multiple monitoring softwares sees CPU temps in the negatives while air cooling in 25 C ambient).
3) Some initial parameter not discussed by the builder was ignored for one build or the other - I didn't see an excerpt for methodology or photographs of the testing setup. Even pros make mistakes.
4) Something I didn't think of in the last three minutes.

Point: There are always more than two possibilities. Knowing what I know (admittedly, less than a hardware engineer and more than the average user), I would also be suspicious of readings that high on hardware that the manufacturer's data sheet lists as having an operational temp range between 0 C and 70 C.

Z

It is clear that you didn't fully read or understand my post, or you didn't watch of fully understand the linked video or some combination of the two.

As well when operational temperatures are listed it is almost always exclusively (I've yet to see otherwise anyhow) in the context of the environment the equipment will be operating in... so your room or more specifically the drive bay can be 70c and the drive will still work. This is why you will see non-operating numbers with an even wider thermal range... This rating has nothing to do with how hot the device can get.

Modern consumer silicon can run at 125c and be within it's thermal limits. It all depends on the actual chip.

There are multiple published test that the show the 941 operating in the range specified in the video.

There are no other possibilities. I was simply letting him figure for himself the correct answer by more carefully considering the question he was asking.

If you wish I will provide links to back my assertions (or any you wish to contest) I will, otherwise I will leave it as an exercise left to the reader to determine the facts.

Point: There is only one possibility when the facts are a known quantity.

Bonus points for contextually ironic Plato quote :P


Personally I though it was true also but was told by the chap who was producing the 3D printed drive cages here that there was no way it would
ever get to this temp via a PM. I just wanted to know one way or the other as I was planning on putting the M.2 overlapping a drive with a max
of 70.c so if it reaches 106.c it's not going to be very good for it's health and / or any plastic printed part one would think unless it's thermoplastic
with a very high temp tolerance.

Aeolisio is good people and very knowledgeable. He is just wrong here if you are accurately representing his statements. It happens to the best of us.

To be clear we are talking about the very hottest part of one chip giving those temps... by dead reckoning the same chip appears at least 15-20c cooler away from the hotspot. The rest of the drive is running much cooler.

Edweird
09-28-2015, 01:08 PM
I can confirm this is true (at least, for myself, I don't have any photos to provide as proof) - these can get very hot under very specific full-on torture tests with benchmarks.
A friend of mine got a 128gb SM951AHCI drive and ran it for several minutes under load and we measured it at 99-102 degrees using a laser termometer. This was only on the controller, though. The rest of the silicon was running considerably cooler.

Under normal work stuff like copying over a few files and such we couldn't get over 45 or so degrees, however.
Still, I would still get some thermal tape and heatsinks - just for the cool factor (heheh) and because they cost a lunch.
Personally, I bought a 250GB 850 EVO with RAPID Mode enabled (which takes tinkering to get working on W10), does the job just as well, it seems. It's the IOPS that really seems to make the difference, not so much the read/write speeds and with RAPID enabled, benchmarks return around a max of 130k IOPS on the admittedly cached data.

Dr. Zchivago
09-28-2015, 01:12 PM
It is clear that you didn't fully read or understand my post, or you didn't watch of fully understand the linked video or some combination of the two.
Your post was straightforward, and solid fallacy. (conjunction, appeal to probability, slippery slope fallacy, possible appeal to tradition)


As well when operational temperatures are listed it is almost always exclusively (I've yet to see otherwise anyhow) in the context of the environment the equipment will be operating in... so your room or more specifically the drive bay can be 70c and the drive will still work. This is why you will see non-operating numbers with an even wider thermal range...

Modern consumer silicon can easily run 125c and be within it's thermal limits. There are multiple published test that the show the 941 operating in the range specified in the video.
Will still work? Probably.
Will likely be reliable for the duration of its warranty period? I'm not the engineer who established the limitations; but, I can say that thermal conditions above and beyond listed norms drastically shorten the lifespan of electronics.


There are no other possibilities.
I listed other possibilities, you are ignoring them.


I was simply letting him figure for himself the correct answer by more carefully considering the question he was asking.

If you wish I will provide links to back my assertions (or any you wish to contest) I will, otherwise I will leave it as an exercise left to the reader to determine the facts.
The facts cannot be inferred from insufficient initial information.


Point: There is only one possibility when the facts are a known quantity.
The facts are not all known; your induction is flawed.


Bonus points for contextually ironic Plato quote :P
In the absence of sufficient support for argument, an individual who immediately resorts to meanness is less likely to continue a reasonable discourse. (red herring)

Z

joshindaphils
09-28-2015, 01:36 PM
Your post was straightforward, and solid fallacy. (conjunction, appeal to probability, slippery slope fallacy, possible appeal to tradition)


Will still work? Probably.
Will likely be reliable for the duration of its warranty period? I'm not the engineer who established the limitations; but, I can say that thermal conditions above and beyond listed norms drastically shorten the lifespan of electronics.


I listed other possibilities, you are ignoring them.


The facts cannot be inferred from insufficient initial information.


The facts are not all known; your induction is flawed.


In the absence of sufficient support for argument, an individual who immediately resorts to meanness is less likely to continue a reasonable discourse. (red herring)

Z

Okay.. I'm not going to break down your entire post that willfully ignores my assertions (and willingness to back them) as well as my clearly stated intent and posits unrelated concerns as a retort (straw man).

Your ignorance on the matter regardless of the materiel here is not an effective rebuttal in particularly given my open invite to back my assertions on request. Not to mention there is the wonderful thing called Google that could be used to educate ones self and shows otherwise.

I ignored your possibilities... because it is clear you have not even reviewed the video or understand it, as you posit things that are clearly not possible given the test shown. As well you show ignorance (operational temp definition) on the subject mater.

Again I am willing to back any of this on request if you have contention with it. Before you do so I suggest researching the matter yourself.

Really just look at the post above yours for confirmation... huh.

Glad your debate class paid off so well.

Sorry you took offence at noting my amusement at your Plato quote it was meant in jest. Just like the debate jab :P

Good day.

joshindaphils
09-28-2015, 01:48 PM
Personally I though it was true also but was told by the chap who was producing the 3D printed drive cages here that there was no way it would
ever get to this temp via a PM. I just wanted to know one way or the other as I was planning on putting the M.2 overlapping a drive with a max
of 70.c so if it reaches 106.c it's not going to be very good for it's health and / or any plastic printed part one would think unless it's thermoplastic
with a very high temp tolerance.

If you are looking to run two drives in the drive bay check out this thread: https://rog.asus.com/forum/showthread.php?58104-How-to-add-a-4th-drive-to-your-G751 There are several people there doing the same with longer length cards you may want to reach out to for feedback. As well there is a link back to the 3d printed bracket thread where others have done the same as well.

I've heard of no one having an issue doing this. If you want to be extra safe add heat sinks, you will be fine.

Dr. Zchivago
09-28-2015, 01:56 PM
Okay.. I'm not going to break down your entire post that willfully ignores my assertions (and willingness to back them) as well as my clearly stated intent and posits unrelated concerns as a retort (straw man).
Straw man fallacy requires me to have refuted an implied argument. I have directly refuted your stated argument, and implied nothing.

Your statement was perfectly clear, and the video shows the temperature of two disks during transfer of equivalent file sets. And, it states clearly the highest temperatures recorded during the process.
In the absence of any other information, your original statement is flawed - you draw a conclusion from insufficient initial information.

Regarding the remainder of your most recent response:
*You claim it is obvious I did not understand the content of the video, yet you do not prove it.
*You say I am ignorant of operational temp definition, yet I never stated the disks are operating beyond limits. I did state, " I'm not the engineer who established the limitations; but, I can say that thermal conditions above and beyond listed norms drastically shorten the lifespan of electronics."
*You state, more than once, a willingness to provide support for your statements; yet, you do not.
*I have never taken a debate class, nor participated in debate, nor have I taken offense to anything. I must, however, point out the flaws in your logic for the benefit of the reader.

To FalloutBoy:
I apologize for joshindaphils and I hijacking your thread.
The recorded temperatures are high; but, without an absolute reference temperature, it cannot be stated that the temperatures recorded are higher than nominal, or are too high.

joshindaphils
09-28-2015, 02:48 PM
Straw man fallacy requires me to have refuted an implied argument. I have directly refuted your stated argument, and implied nothing.

Your interpretation.



Your statement was perfectly clear, and the video shows the temperature of two disks during transfer of equivalent file sets. And, it states clearly the highest temperatures recorded during the process.
In the absence of any other information, your original statement is flawed - you draw a conclusion from insufficient initial information.



The absence of additional information is a false construct, information abounds, you just choose to ignore it... for the sake of argument perhaps? I'll demonstrate later in this post.



Regarding the remainder of your most recent response:
*You claim it is obvious I did not understand the content of the video, yet you do not prove it.


Your two points here:

1) Assuming the same system for both drives, there exists the possibility of a faulty temp sensor.
2) The monitoring software could be in conflict with another process (as when a system with multiple monitoring softwares sees CPU temps in the negatives while air cooling in 25 C ambient).

Are nonsense as it is clearly video from a FLIR camera being displayed on the video 1) could be bent to fit 2) no way.



*You say I am ignorant of operational temp definition, yet I never stated the disks are operating beyond limits. I did state, " I'm not the engineer who established the limitations; but, I can say that thermal conditions above and beyond listed norms drastically shorten the lifespan of electronics."


Your statement: I would also be suspicious of readings that high on hardware that the manufacturer's data sheet lists as having an operational temp range between 0 C and 70 C.

This clearly implied to me that you expect the drive to not exceed these temps, though rereading it I can see another, abet still misguided interpretation where the environmental and max safe device temperature are directly tied.

From Wikipidia:

Devices are manufactured in several grades. For example, the manufacturer Altera defines five temperature grades for its products, each with an associated normal operating temperature range:[1]

Commercial: 0C to 85C
Industrial: −40C to 100C
Automotive: −40C to 125C
Extended: −40C to 125C
Military: −55C to 125C

These grades ensure that a device is suitable for its application, and may withstand the environmental conditions in which it is used.

Could not find exact specs for the 941, but the below link has the data for the 951 and clearly marks the operation temp under environmental stats:

http://www.tweaktown.com/reviews/6951/samsung-sm951-512gb-2-pcie-ssd-review/index.html



*You state, more than once, a willingness to provide support for your statements; yet, you do not.


It was very clearly stated that this would be done on request or in a similar vain each time. He you actually provided direct concerns and here you have my support.



To FalloutBoy:
I apologize for joshindaphils and I hijacking your thread.
The recorded temperatures are high; but, without an absolute reference temperature, it cannot be stated that the temperatures recorded are higher than nominal, or are too high.

Here is another additional data point that shows similar temps (scroll down a bit)

http://www.tweaktown.com/reviews/6372/samsung-xp941-512gb-m-2-pcie-ssd-review/index10.html

Here is a third:

https://rog.asus.com/forum/showthread.php?77302-Can-M-2-SSD-s-really-get-this-hot-or-is-it-fake&p=540433&viewfull=1#post540433

Please feel free to Google for more confirmations.... this is a very popular drive for being only OEM marketed, lots of reviews and information out there.

Dr. Zchivago
09-28-2015, 04:45 PM
Your interpretation.
If that is the case, then I require your definition for straw man.


The absence of additional information is a false construct, information abounds, you just choose to ignore it
I must add: for some reason I have yet to explain, the video notes (immediately below the video player, including publish date, that methodology I mentioned earlier), were not present when viewed on my desktop, earlier. After a new review on my laptop, I see clearly all of the information I was looking for. I am now 'suspicious,' because Firefox (and all of the addons) on my laptop shows that section, and Chrome (including all of my addons) on my desktop does not.
Therefore, I must rescind earlier statements regarding earlier statements about 'lacking information.'

However, I said that you didn't provide necessary support for your initial claim; I did not state additional information doesn't exist. From only the premises you provided initially (including that which is available elsewhere), your original conclusion was false. That is still true.


Your two points here:
1) Assuming the same system for both drives, there exists the possibility of a faulty temp sensor.
2) The monitoring software could be in conflict with another process (as when a system with multiple monitoring softwares sees CPU temps in the negatives while air cooling in 25 C ambient).

Are nonsense as it is clearly video from a FLIR camera being displayed on the video 1) could be bent to fit 2) no way.
1. Specifically, temp sensors - not method of installation. The possibility exists for a physical hardware flaw in the sensor(s).


Your statement: I would also be suspicious of readings that high on hardware that the manufacturer's data sheet lists as having an operational temp range between 0 C and 70 C.

This clearly implied to me that you expect the drive to not exceed these temps, though rereading it I can see another, abet still misguided interpretation where the environmental and max safe device temperature are directly tied.

It wasn't my intention to suggest that the suspicion is toward the reviewer - the suspicion is directed at the drive itself. In fact, "suspicion" was a poor word choice on my part (which, resulted in misinterpretation). Perhaps, the statement would be better worded: "I too would wonder if this drive should run this hot during a simple file transfer (albeit, for a short period - and, I did note that the drive cooled very quickly); and, I wonder how much hotter it will become during heavy use, and how will that affect my other components? How does this drive compare with other manufacturers, etc?"
The point here being: the original question was the result of just such suspicion.

Z

EDIT: I forgot to mention: I'm going to install and sync Chrome on my laptop to try to recreate the aforementioned issue.

joshindaphils
09-28-2015, 05:15 PM
Another day or verbally I'd love to continue the conversation, but I need to hang this up today.

Regarding straw man, your interpretation of your statements. Not the definition of straw man... Intent and reception are often different.

Here is my view on the events that unfold (forgive the liberal license taken)

OP: Hey are these temps real or fake?

ME: Given this what do you think?

OP: Yeah I figured they were real.

You: No way you are not considering all this stuff that could be going on
(which I still don't understand your explanation of BTW)

Me: eh I hear you, but the derived answer is correct I can show you if you want.

You: oh but this and that nothing to back my insistence that the accuracy of the conclusion is pure conjecture and ignoring your suggestion to look into the matter myself and your offers to back your own assertions

Me: yeah see look here it is accurate okay

You: ah yeah I missed that, but other stuff

---

Look I work simply for the most part, though at times will posit answers as questions to engage a bit of additional thought on behalf of the person initially asking. I am not writing a proof I need not be wholly inclusive of every conceivable explication when I know the facts and the answer is correctly derived.

You could just as easily tear apart or provide theories against why the sky is blue... it doesn't help anyone.

Q: Are these numbers accurate?
A: Yes they are.

QED

aeolisio
09-28-2015, 10:47 PM
First off the hottest part of the m.2 faces up not down towards any second drive in the same bay. The underside does not get anywhere near the temp of the top use a physical probe and check for yourself.
Secondly i recommended heatsinks and was very clear in real world everyday usage you would not see that temp. A stress test or benchmark is not anywhere near the same as file transfers or running a game. In 90 percent of uses the sm951 will not be the bottleneck in the system. The sm951 also runs considerably cooler then the xp941. We have run much heavier loads then most users on over 30 sm951s at the office and never peaked above 70. Now if you take a naked xp941 stick it on a desktop motherboard right next to a cpu with whatever ambient they felt like using then thats not really the same thing is it. If fallout already has a m.2 he can just pick up a temp probe and see what temps he actually get on the underside of the drive himself with his workload and use that to make his decision. Thats what i did when i was deciding if i wanted to even make a variation of the bracket for dual drives in 1 bay. But please dont waste my time with multiple messages and then make a thread about it. The only reason i even answer messages is to help people avoid new threads.