If only they'd use something other than Flash. There are many better technologies
which could be used, as reported in New Scientist and Scientific American over the
years. Heck, there was a pseudo solid state device developed 15 years ago (the
"Trillion Bit Cube", ie. 116GB; New Scientist, 13/Aug/1994, Issue 1938, pp. 23-23)
which offered more than 10X faster sequential I/O than ST's 2TB unit, though the
access time was not as good. Made from an alcohol-based polymer, one laser
illuminated a slice through the block, another illuminated a point on the slice.
No moving parts for the recording medium, but of course some moving parts were
required to control the lasers, thus the lesser access time to find the required
X/Y/Z, but once at the right location the read/write rate was very fast. At the
time, the test device was said to be about the size of a VHS tape.
Spintronics, memristors, quantum devices, all sorts of things in the pipeline.
Remember folks, the current products using Flash RAM are just the beginning, and
compared to what ought to be possible they're actually very slow (presumably the
fastest ever possible speed would be 10^34 bits/sec, ie. the limit of quantum
events). Alas, as always it's about commercial realities, getting products to the
market, earning revenue, developing the next gen product, making a profit to
satisfy shareholders, weathering economic ups & downs in the meantime, and the
individual staff involved just trying to survive, etc. Just wish for once IBM or
whoever would go for one giant leap all in a single step instead of drip-fed
small-fry improvements.
Ian.