I am just guessing: Could sector size have something to do with it?

And no, I wouldn’t trust it.

So, would you trust this drive?

No

Hell no, that kind of behavior is what those aliexpress/wish knockoff drives do to fool the customer base stupid enough to trust the listing.

I would go so far as to say to get a refund and go to an actual store so you’re not buying bootlegs.

As others said, this looks like a fake drive sold as a scam. Chances are high the drive controller has been modified to report incorrect values. Any data written to that space beyond its actual capacity will either overwrite existing data or go straight to dev/null

There are SSD and usb stick drives with a fake size programmed into it. There are scams who sell higher capacity, but delivering lower capacity. Looking in file browser (or other tools), it still looks like 1 TB. But as soon as you fill the real 64 GB, the rest of the data is written into the void = data loss. So from the looks like this drive was originally a fake drive. And whoever sells it to you probably knows it and tries to sell it correctly as a 64 GB (or maybe the original size is even different), after getting scammed maybe?

Whatever the original story is, this drive looks fishy, acts fishy and is probably a fish. I don’t know how much you spend on this, but I would not use it, throw it away. Please don’t give or sell it to someone else, or keep it as an evidence. If possible, report the person who sold it to you.

I’d test it with a program like https://github.com/AltraMayor/f3

If you want to test it yourself, you could use a tool like f3:

f3 is a simple tool that tests flash cards capacity and performance to see if they live up to claimed specifications. It fills the device with pseudorandom data and then checks if it returns the same on reading.

…readthedocs.io/…/introduction.html

As long as whatever software you’re using can handle a cache failure, I can’t see why not. I’d put my ~/.cache on there no problem.

If you can return it, return it.

If not, nuke it: overwrite the first MB or so with dd. You could also tell gparted to create a new partition table. Then, reboot, and try again to see what it reports.

And no, I would not trust it to do crucial stuff in any case. And 64GB isn’t much anyhow.

To test, write known random data to fill the drive at the block device level, reboot or otherwise clear the buffers, and then read the whole drive again validating the data is exactly correct. These fake drives should not be able to fake random data filling.

Are you super sure you’ve selected the correct drive in gparted?