[SOLVED] M.2 SSD Problems

ccrocker58

Prominent
Mar 16, 2020
17
1
515
A couple weeks back I had a Samsung EVO 970 that would, during extended periods of read/write, just flat out stop working. If I was playing a game that was installed to the drive, the game would immediately crash to desktop, and anything that was downloading would immediately fail. Trying to access the drive would sometimes let me see the contents of the drive, but I couldn't actually run anything (I could see I had a game installed, but could not run it). I assume this was only possible when I had windows displaying the contents of the drive open prior to the crash. Restarting my PC would fix this problem, letting me access the SSD again, at least until the issue popped up again.

With four days left on my return policy, I decided to just get a refund for it on the chance that it may have been a faulty SSD, and picked up a 970 Plus. Unbelievably, while I was waiting for it to ship to me, my HDD with Windows 10 installed on it died, so I figured I would install it on my incoming SSD.

I'm fairly certain the same problem is occurring now as Windows will randomly just freeze. So I'm pretty sure I can rule out the SSD being the problem. Does anyone have any idea what may be going on here? At this point I'm thinking there's something wrong with my mobo; I have the SSD in the m.2_1 slot on my rog strix B450-F gaming board since the 2_2 slot would throttle the PCIe x16 slot down to x8 mode. The only other thing I can imagine being a possible culprit would be the temperature. This thing is running very hot, at idle it's around 45-48c, during extended write sessions it goes up to 54, and during gaming I've seen it reach 59.
 
Solution
I don't think you are reading your motherboard manual correctly. Locating your SSD in socket M.2_2 will not reduce the speed of your first PCIe 3.0 x16 slot down to x8. That is only if you are using an APU like the Ryzen 3 2200g. You can also mount your graphics card in the 2nd x16 slot. It runs at the same speed as the 1st, PCIe 3.0 x16.

Is your setup using 3 sticks of RAM? If yes, you should probably just use the 2 sticks of GEIL and leave out the 1 stick of Crucial. Using 3 sticks of RAM will force the motherboard to run in single channel mode instead of dual channel.
Mar 9, 2020
18
1
20
Overheating is probably the cause since it happens after a certain period of time elapsed when in use. The temperature you describe at idle is way to hot !
Around 30-35 degrees celcius should be considered average when idle.
If i were you i'd try cooling the M2.SSD by adding a cooling-sink or extra case-fan that targets the SSD.
 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Those temperatures are not unusually hot for an M.2 SSD. Have you listed all the parts in your PC?

I have not yet, but will as soon as I get back from work. After you mentioned that, I realized that it may actually be an underpowered PSU. A couple months back I swapped in a GTX 1070 which was a substantial upgrade and upgraded my mobo as well. I suppose that, combined with my AIO and two HDDs may be drawing more power than it can handle.
 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
A little disclaimer before I go on; I am relatively new to PC building as I only recently decided that I wanted to shift my gaming over to it. So I'm sure there's a lot I don't know. This particular PC was one that was given to me after a friend built a new one. I have swapped a lot of the hardware, but for some things that came with it I'm not positive on.

Anyway, my setup:

Mobo: Rog Strix B450-F Gaming
CPU: Ryzen 5 1600X
GPU: GeForce GTX 1070 (seated in PCIe 3.0 x16_1 slot)
RAM: 1x Crucial 4gb (not sure of the model, but it's extremely basic looking; no plate on it or anything) 2x GEIL Super Luce 4gb

Nothing is overclocked.

I'm not sure if it's necessary to know, but I did just swap out the PSU that was with it when it was given to me (600w) with mine (750w). As far as cooling goes, the CPU is cooled by a Cooler Master AIO (unsure of the model as it was given with the PC) while the case itself has 4 fans. As mentioned, the 500gb Samsung EVO 970 Plus is seated in the M.2_1 slot currently. This slot is unfortunately located directly above my GPU, so it receives a lot of ambient heat from it. Even if the aforementioned temps are normal for a SSD, I would be interested in anything that could drop those numbers a bit, but I'm unsure if there's anything I can do given the very limited space at the location of the slot it's in. I'm not entirely opposed to moving the SSD down to the M.2_2 slot; I've done some research and it seems dropping the PCIe 3.0 slot down to x8 mode would only make a difference if I were trying to run games at higher FPS and higher resolutions, and even then it seems the drop is only about 1% in most cases. I'm not looking for anything more than 1080p/60FPS, so I think I would be fine.

I truly appreciate any help that can be given, and if there's any more info that would be needed and/or helpful, please do let me know (that hardware list seems really small compared to some of the huge lists I've seen in other posts, so I'm sure I'm missing stuff there that would be important)!
 
I don't think you are reading your motherboard manual correctly. Locating your SSD in socket M.2_2 will not reduce the speed of your first PCIe 3.0 x16 slot down to x8. That is only if you are using an APU like the Ryzen 3 2200g. You can also mount your graphics card in the 2nd x16 slot. It runs at the same speed as the 1st, PCIe 3.0 x16.

Is your setup using 3 sticks of RAM? If yes, you should probably just use the 2 sticks of GEIL and leave out the 1 stick of Crucial. Using 3 sticks of RAM will force the motherboard to run in single channel mode instead of dual channel.
 
Solution

ccrocker58

Prominent
Mar 16, 2020
17
1
515
I don't think you are reading your motherboard manual correctly. Locating your SSD in socket M.2_2 will not reduce the speed of your first PCIe 3.0 x16 slot down to x8. That is only if you are using an APU like the Ryzen 3 2200g. You can also mount your graphics card in the 2nd x16 slot. It runs at the same speed as the 1st, PCIe 3.0 x16.

Is your setup using 3 sticks of RAM? If yes, you should probably just use the 2 sticks of GEIL and leave out the 1 stick of Crucial. Using 3 sticks of RAM will force the motherboard to run in single channel mode instead of dual channel.

I see. Yeah, I guess I was misreading my board's manual in that case. I also will remove the Crucial; I planned on getting another 2 sticks anyway.

That said, after posting my last message, I restarted so Windows could install the remaining updates it needed, and it once again froze up while it was installing the update. I've moved the SSD to the M.2_2 slot now and removed the Crucial stick. It's currently re-downloading the failed update...but the temperature on this thing is now the highest I've seen...well anything in any of my PC's as it's currently sitting at 61c as this update is in progress. Hard Disk Sentinel seems kind of concerned with red warning signs, and Samsung's own Magician straight saying it's "Too Hot". Are these temps really normal?
 
What’s A Safe Operating Temperature For SSDs?

Most SSDs are rated for running within a temperature range of 0ºC up to a max temp of 70ºC (32ºF to 158ºF). This range is about the same for every consumer SSD currently on the market, but can sometimes vary slightly based on the model and form factor of the SSD (2.5” SATA, M.2 or mSATA) – but a drive under 70ºC is generally within the operating parameters provided by SSD manufacturers.

Samsung SSD Temps: 950, 960, 970, 750, 840, 850 and 860 EVO and PRO series are rated for operation between 0ºC and 70ºC.

 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Mkay, so I can rule out the temperature being at fault here.

After downloading that update, this time it did install it without any problems. One thing that I did notice is that when I first removed the Crucial stick, I left the remaining geil sticks in the same lanes, which was not the recommended configuration (A2 and B2 is the recommended; the sticks were in A2 and B1, with the removed Crucial in A1) and when it got to my desktop it was an immediate lock. Hopefully the problem here was an improper RAM configuration, though I would be curious to know why this configuration seemingly worked fine for everything but messed with the SSD.
 
Last edited:
Mkay, so I can rule out the temperature being at fault here.

After downloading that update, this time it did install it without any problems. One thing that I did notice is that when I first removed the Crucial stick, I left the remaining geil sticks in the same lanes, which was not the recommended configuration (A2 and B2 is the recommended; the sticks were in A2 and B1, with the removed Crucial in A1) and when it got to my desktop it was an immediate lock. Hopefully the problem here was an improper RAM configuration, though I would be curious to know why this configuration seemingly worked fine for everything but messed with the SSD.
That could possibly explain your problem. If the sticks were in A2 and B1 they were not working together, they were in different channels. That would also mean that 1 stick of GEIL and 1 stick of Crucial memory were in the same channel which could cause problems.
 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Sigh...Windows downloaded the final round of updates as I was typing my last message, and after posting I restarted the PC. It got to the login screen and locked again...frustrating...
 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Device Manager did have an error I've never seen before during a clean install this time around, but it was with my GPU. Basically, Windows took a very long time to actually recognize my GPU at all. When I first installed Win10 on the SSD, the device listed under Display Adapters was listed as a Windows device of some kind, I don't recall exactly what it was called, with an error sign under an "Unkown Device" which listed "PCI Express Controller". This had never happened before. After about 10 minutes everything seemed to fix itself and my GPU was there again. I assume it took a longer than usual time to download the NVIDIA drivers it needed.

I'm honestly not sure if the mobo drivers are up to date...would it be the ACPI x64 Based PC listed in the Device Manager? If so, right clicking and selecting Update Drivers comes back saying it's up to date.

I can definitely say that I haven't updated the BIOS since I got the mobo. I'll try doing so.
 

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Working on downloading everything now. It will likely take me a bit due to only having wifi atm.

In the meantime, I would really like to thank you for all your help and advice. Regardless of whether or not this gets fixed, I really appreciate your assistance and patience.
 
  • Like
Reactions: rocktalkrock

ccrocker58

Prominent
Mar 16, 2020
17
1
515
Have you looked in Device Manager for any yellow warning signs?
Are all your motherboard drivers up to date?
Does your motherboard have the latest BIOS update?

Well, I'm cautiously happy to report that it seems the issue has been resolved.

I went ahead and updated the BIOS, audio driver, ethernet driver, and the chipset. I ignored the Utilities section as that seemed irrelevant, and the RAID drivers as I don't plan on using it any time soon. While downloading the drivers, I also went ahead and did a little more research on the issue and found a thread on the Samsung forums that detailed a problem very similar to mine. On there it was noted that the base 970 EVO had problems hanging and freezing in Windows when, ironically, Samsung's own drivers were used for the drive. I went ahead and also uninstalled the Samsung drivers. I've been using the PC since my last post without interruption while downloading a 17gb game, playing a game which is installed on the SSD, watched a few gameplay recordings I need to edit, etc. Basically, I just did what I would normally do throughout the average day.

During all this the temperature also was between 41-49c, which was nice.

That said, that was all pretty light work. I guess I'll know if everything is truly okay after I try to do something more taxing on the drive. Is there a stress test that you would recommend for this purpose? So far, downloading and installing Windows updates seems to have been the thing that's pushed this drive the hardest judging by the fact that the temperature was hitting it's highest point during them, but with no more updates available I can't exactly use that as a test anymore. I know Magician has its benchmark, but that completed with no problems the whole time I have been dealing with this, and didn't really seem to push the drive too, too hard.
 
I just use the Samsung Magician software tool to test my drives but you can also use CrystalDiskMark. I also like using a small program called Speccy to monitor all the different parts in my system, especially the temperatures.

installed-hardware.png