Is NVMe M.2 expected to run relatively hot?

CountMike · 07 Apr 2021

jimbo45 said:

Hi there
basic physics

Current is conducted by the movement of electrons through an object -- the more atoms that are crammed together in an object (inevitable if you have larger and larger capacity devices in smaller footprints) the more heat will be created by the movement of electrons through the device -- offset partially by the lower current needed in the first place - so there's always a trade off between "minaturisation" and heat generation.

They shouldn't though run hot (as opposed to a bit of warming) - nothing wrong with heat sinking a device if it needs it but if this sort of device is running hot I'd look for other problems - e.g lack of adequate cooling for the whole box.

Cheers
jimbo

High frequency can also produce heat (think microwave). With SSDs it's controller chip that gets hottest.

jimbo45 · 07 Apr 2021

CountMike said:

High frequency can also produce heat (think microwave). With SSDs it's controller chip that gets hottest.

Good point - although not sure if SSD's use HF / microwaves -- I wouldn't put a damp SSD device in a microwave oven to dry it out if it got wet !!!! - It might survive though but I'm not going to perform that experiment. !

Cheers
jimbo

CountMike · 07 Apr 2021

jimbo45 said:

Good point - although not sure if SSD's use HF / microwaves -- I wouldn't put a damp SSD device in a microwave oven to dry it out if it got wet !!!! - It might survive though but I'm not going to perform that experiment. !

Cheers
jimbo

PCIe 3 clock frequency is 8GHz., well into microwave spectrum.

jimbo45 · 07 Apr 2021

CountMike said:

PCIe 3 clock frequency is 8GHz., well into microwave spectrum.

Cheers

we learn something every day (but I'm still not putting a wet SSD into the Microwave) !!!!!!

jimbo

Helmut · 07 Apr 2021

Yes, they are designed/optimised to run around the 60°C mark, particularly the high speed drives like those Samsung ones.
They will throttle down if they get too hot, part of the SSD controllers job.

It is not a given that lower temperatures are better, in many scenarios the failure rate is higher.
You can look up the data if you wish.

It is all a complex compromise between Speed, density of components, thermal considerations and a sensible range of operational temperatures.

"High frequency can also produce heat (think microwave). "
Just no, microwave ovens (heating) is a bad comparison.

ptrthgr8 · 07 Apr 2021

Thank you for all the great info, everyone! I feel better knowing that what I'm seeing isn't really unexpected, although I do think I'll add an intake fan on this case to help with cooling in general. And I'll start using HWMonitor instead of Speedfan. :)

Thanks again and cheers!