Not to shill for a company, but probably because it's rated to work with it. Similar experience you get with "enterprise equipment".
"You could use any other drive, but our drive has been through rigorous testing" kind of situation.
Ideally we wouldn't need this since standards exist, but time and time again somebody is bound to take a shortcut and break things. Be it Raspberry in their PCI-E implementation or the drive manufacturer skipping a few NVME functions to save few kilobytes of firmware.
Think of it as a guaranteed "trouble-free" experience if you just want to plug it in and work.
I have been boiled by these cheap SSDs once, and it was a firmware related issue too.
200USD for half-assed 4TBs of SSD storage that may or may not work depending on what you plug it into.
PS: It was a Silicon Power SSD as well, so really do watch out for that stuff.
I see you've never had the awfulness of dealing with the weirdly-keyed SSDs or the SATA SSDs! They're terrible! I had one of those laying around in an older mobo and I wanted to just put it in another machine, nope. I found an SSD that's the same size, got an M.2 to USB adapter that handles all of them, then just used dd to copy the data over (since they're the same size, you can just dd it directly). Not fun.
Given that banana countries were a thing, I think it wouldn't be completely out of character for the US to enforce it's (it's own companies) interests. But I don't think it will be war, but rather soft-power.
To be fair to them, some Linux defaults are atrocious IMHO.
For example mq-deadline ALWAYS performs badly in desktop tasks. Launching a Steam download alone is enough to kill a Linux machine using that scheduler. Therefore I feel like we need this.
Unfortunately it tends to do so randomly (was using foot for like 3 or so years by now). And in the end you'll only ever end up saving single digits of RAM. I suggest using server mode *only* on a RAM constrained device like a SBC or a embedded device (Tablet running Linux).
"You could use any other drive, but our drive has been through rigorous testing" kind of situation.
Ideally we wouldn't need this since standards exist, but time and time again somebody is bound to take a shortcut and break things. Be it Raspberry in their PCI-E implementation or the drive manufacturer skipping a few NVME functions to save few kilobytes of firmware.
Think of it as a guaranteed "trouble-free" experience if you just want to plug it in and work.
I have been boiled by these cheap SSDs once, and it was a firmware related issue too.
200USD for half-assed 4TBs of SSD storage that may or may not work depending on what you plug it into.
PS: It was a Silicon Power SSD as well, so really do watch out for that stuff.