[SOLVED] SSD suddenly overheats and dies

[mikael_schiffer]

I bought a Adata SX 6000 512GB NVME SSD from from an online store recently
On installation it started having problems. Sometimes, if stressed a bit,like copying files for sometime, it overheats and simply dies.
Using Adata's SSD Toolbox software, i monitored it. When the SSD utilization hits 100% , the SSD Toolbox crashes. When i reopen the SSD Toolbox app, the temperature shows 60000+ degrees, and the drive becomes inaccessible (as shown in the screenshot)
I have to restart my PC to get make the SSD show up again.

However, sometimes it works absolutely fine. I even managed to install games and run some Crystal mark benchmarks.
But out of nowhere the overheating problem occurs again.


My specs:

Amd Ryzen 1500x
8GB Ram
Gigabyte B350M D3H
Gtx 1050ti
Antec VP 500
other storage: 1 TB HDD and 120GB SSD (both SATA)
Windows 10

Pleaze help i spent a lot of money on this
Is this some software related issue (which i can fix) ?

 

[cin19]

You should contact adata for replacement.

 

[CountMike]

According to SSD Toolbox it's running several times hotter than sun !!!! Something is screwed up in it's firmware.

 

[USAFRet]

"I bought a Adata SX 6000 512GB NVME SSD from from an online store recently... "

Warranty replacement.
Or preferably, a full refund, and buy a better drive.

 

[fzabkar]

65263 = 0xFEEF (a negative hexadecimal number)

0x10000 - 65,263 = 273

So the SSD is actually reporting a temperature of -273 degC (= 0 degrees Kelvin = absolute zero).

In other words, it appears that the SSD does not have a temperature sensor.

You can view the raw SMART attribute data using CrystalDiskInfo:
https://crystalmark.info/en/download/

CrystalDiskInfo will probably report a raw value of 0xFEEF for the temperature. Adata's tool interprets this as an unsigned 16-bit number, so it evaluates it as 65263 in decimal. The SSD firmware, OTOH, understands this to be a signed 16-bit number, in which case it evaluates to -273.

 

[mikael_schiffer]

"I bought a Adata SX 6000 512GB NVME SSD from from an online store recently... "

Warranty replacement.
Or preferably, a full refund, and buy a better drive.

The online store says i have to RMA it.
I contacted ADATA and they said i have to go to their service centre for RMA. There is no service centre in my State and ADATA does not accept RMA via courier. I have to go there personally.

 

[mikael_schiffer]

65263 = 0xFEEF (a negative hexadecimal number)

0x10000 - 65,263 = 273

So the SSD is actually reporting a temperature of -273 degC (= 0 degrees Kelvin = absolute zero).

In other words, it appears that the SSD does not have a temperature sensor.

You can view the raw SMART attribute data using CrystalDiskInfo:
https://crystalmark.info/en/download/

CrystalDiskInfo will probably report a raw value of 0xFEEF for the temperature. Adata's tool interprets this as an unsigned 16-bit number, so it evaluates it as 65263 in decimal. The SSD firmware, OTOH, understands this to be a signed 16-bit number, in which case it evaluates to -273.

After i shut down the PC and i touch the SSD , its not very hot. Maybe the SSD thinks its running too hot because its sensors are bugged.
Can i fool the SSD to think its not running hot? Maybe limit SSD utilisation to 90% maximum. I wouldnt mind slowing the SSD down to 100Mbps. As long as it is usable i am fine.

 

[cin19]

May try to update the AMD chipset driver, check your SSD with toolbox to see there is the new firmware to update or not.
AMD chipset: https://www.amd.com/en/support

You can't fool the SSD or the PC, but you can ignore that temp, and use the SSD until it is dead if you want.

 

[fzabkar]

The SSD doesn't need to be fooled. It understands its own temperature report. The problem is that Adata's tool is unaware of how to interpret the report.

 

[fzabkar]

The SMART report in this review shows a sensible result for the Composite Temperature:

https://www.legitreviews.com/adata-xpg-sx6000-pro-512gb-ssd-review_212291/2

Could the drive be a fake?