Power Cycle a SSD

Sounds unlikely.

It is just possible. It does seem that the interface is bricked rather than the memory itself is bad

Even if it worked, would you ever trust it again.

I had a Samsung EVO fail, intermittently, it got worse over time. Diagnostics tests reported it was fine. I also used it in multiple machines and different OSs, and it eventually went bad in all of them. I should have binned it at the first error.

That's rather how I feel I must say.

Indeed. You should simply return it and ask for a like for like replacement and say that if it is in fact fine then they can resell it.

I have a Panasonic TV bought from Currys. It's predecessor worked fine except if you plugged it into the Ethernet, when it shut itself down.

I took it back and the techTurd tested it and said there was nothing wrong. I pointed out he hadn't plugged it into Ethernet. He said he hadn't got any. I said 'why would I bring a perfectly good TV in for like for like replacement if it wasn't faulty' and he said 'fair point' and I got another one, which has been 100% fine ever since.

This is not your problem. It is Crucials. Send the mother back

It's a warranty not SOGA so I am a bit bound by their process.

Anyway they have agreed to RMA it I am glad to say. Mind you when WD did that for a couple of SSDs neither of the worked, they must have specials for RMA.

Have you said a prayer for the deceased, while it lays on a golden altar ?

Remember, consumer SSDs don't have a SuperCap and don't have "advanced-power-fail" that comes from such power backup schemes.

Anandtech used to show pictures of SSD PCB, where the pads were empty where the SuperCap was located, but also the SMPS boost converter components next to the SuperCap were missing. If this is fitted, it gives the device extra time to put its toys away neatly.

It depends on the partition type, but if it was NTFS for example, you could connect it to a Windows PC and deliver a TRIM to it from the Optimize option. This would inform the drive, which sectors were free to be placed back into the free pool.

But the metadata for the drive, must be stored in some SLC-thresholded area, which allows more write cycles.

The map for the device can be in two places. For a higher priced unit, it comes with a DRAM chip and the virtual to physical map is held in DRAM. For the host buffered SSDs, the map is in host memory, and then there needs to be a scheme so the drive can fetch and use portions of that. And the drive must also ensure the map is never lost.

I don't know why, but a lot of NAND based devices and their controllers, seem to get into a Spiral of Death, where they rush to replace every section of flash, with a "spare". Until there are no spares left. One or two of the Samsung NVMe got into that mode. A firmware update was supposed to stop it... but any spared out areas were spared out for good. The firmware flash did not reset bad block detection. The firmware flash was good for new-in-box items of those model numbers, you unpacked them and flashed them immediately, yielding a usable storage device.

If I had an SSD go like that (go like some of my USB sticks died), I doubt my prayers would do a damn thing for the item.

I suppose if you power cycled it a thousand or even ten thousand times, it would revert to the brand-new state ??? Applying my teenage logic to my purchase "We could also take it out back and set it on fire". Too bad they didn't have magnesium cases.

Paul