I was an SSD fan ...
You can thank me later
? ? ?
What was your involvement?
It's not that opaque. After all, I've been accused of poor opsec.
I always saw storage as the main performance bottleneck in any computer system.
Yup. Disk latencies have up till recently hidden universes of poor design in other layers of the storage stack, OSs, middleware, and applications themselves. With the widespread availability of NAND's much lower latencies, a more efficient protocol than ATA or SCSI -- namely NVMe (though SCSI's SOP/PQI would have been in my opinion a superior technology ((though it's hard to fight Intel))* ) -- is necessary to unlock it's native performance.
*::hrumph!:: what's another borg-ian windmill against which to tilt?
However, NVMe itself is really not efficient enough for NAND. SW stack latencies are still greater than the underlying media's access times. And with a new generation of Storage Class Memory -- another order of magnitude faster than NAND -- just around the corner, this problem will only become proportionally larger.
I figured when cheap low-power-consumption high-capacity dependable solid state storage was achieved, we'd see a whole new world of portable devices.
Even further. As recognition of the inefficiencies that have allowed to fester for a half-century become more widespread, attention will turn to the entire storage stack, the POSIX interface, and applications, ushering in a complete rearchitecting of canonical computer system architecture.