0

I have a Samsung PM991a NVMe SSD, and during idle it's pretty cool (around room-temperature).

However, during filecopy (10ish GB total size), it gets pretty warm (ie. HOT):

enter image description here

Is 100 °C considered too hot for an SSD? I don't mind having it, BUT I want my data safe in case of this temperatures when I'm copying files for longer periods of time (this is 1TB size, so I can copy files for half an hour or even 1 full hour)?

So in a short simple question: can an NVMe SSD survive 1 full hour at 100 °C without data-loss, or it's way above its limits so I can only have 50-50 chance?

Daniel
  • 487
  • 4
  • 15
  • ~65°C is about the limit, you ought to get some air passing over that. – spikey_richie Apr 19 '23 at 19:32
  • 2
    Does this answer your question? [Should I worry about high ssd temperature?](https://superuser.com/questions/1592187/should-i-worry-about-high-ssd-temperature) – spikey_richie Apr 19 '23 at 19:33
  • You should look at the quality of the SSD and the environment you put it in. I have a Lenovo Desktop with a 1 TB Samsung SSD main C: Drive and is just a circuit board. It is running about 28 Degrees C. The second drive is a Samsung 2 TB drive in an aluminum enclosure. I set up a 30 GB copy to it and it stayed steady at 22 Degrees C. Direct temperature readings with an infrared thermometer meter. – John Apr 19 '23 at 20:13

2 Answers2

1

It's not entirely suprising, writing is a particularly demanding process where a high voltage injects electrons into an isolated gate within the silicon substrate.

Whether it is "bad" depends exactly what that specific area of chip is doing. If it is the charge pump PSU inside the chip then it's probably fine, similar for the controller. The NAND itself would be better off cooler.

It's difficult to know exactly what that bit does without dissolving the plastic away and examining the circuitry in detail but if it works then it works.

A heatsink may help, at least keeping performance up. Many faster SSDs do encourage their use as they may start throttling at high temperatures.

Mokubai
  • 89,133
  • 25
  • 207
  • 233
1

The article Samsung PM991 M.2 NVMe PCIe 3.0 SSDs has this specification :

Temperature : Operating 0°C to 70°C Non-Operating -40°C to 85°C

The answer is then that your SSD is exceeding its maximum temperature range, which is not at all recommended.

The question of how long this SSD will be able to function until breakdown cannot be answered. I suggest not finding this out empirically...

You should look at the airways in your computer and clean them well, because such temperatures are very rare, so something is wrong.

Another article from EaseUS, a reputable company, SSD Temperature Range: Everything You Should Know adds this :

Simple read/write jobs do not significantly raise the temperature. So, unless the SSD temperature exceeds 70°C, it normally isn't a cause for alarm.

Your temperature much exceeds this limit.

harrymc
  • 455,459
  • 31
  • 526
  • 924
  • Listed temperatures are often "local ambient" not actual operating tempteratures. They tend to specify that with the area around the electronics at a particular temperature the device may get hotter without killing themselves. It doesn't really tell you what temperature the device itself may get up to. An Intel CPU may have an "operating temperature" of 85C, but the junction temperature where the silicon may be damaged is actually 105C. – Mokubai Apr 19 '23 at 19:35
  • @Mokubai: I added another source that says the same. An SSD doesn't have moving parts, so it shouldn't reach such high temperatures under normal conditions. It's also possible that the disk is defective. – harrymc Apr 19 '23 at 19:41
  • @Daniel: If the controller is the hot part, there might be an electrical problem. I suggest great caution. – harrymc Apr 19 '23 at 19:46
  • `An SSD doesn't have moving parts, so it shouldn't reach such high temperatures under normal conditions.` That's a really funny sentence to read. Sounds like CPU heatsinks are scam. (Not that I think the issue in the OP is fine / "normal" though.) – Tom Yan Apr 19 '23 at 20:30
  • @TomYan: Under normal conditions. – harrymc Apr 19 '23 at 20:37
  • "`moving parts`" – Tom Yan Apr 19 '23 at 20:39