PC Perspective’s Allyn Malventano stopped in earlier this week and sent me a link to his take on the bleak future of SSDs and flash memory: http://www.pcper.com/reviews/Editorial/NAND-Flash-Memory-Future-Not-So-Bleak-After-All
He didn’t agree with me entirely–he argued that the problems outlined in the study in question are solvable.
Mr. Maltevano clearly knows more about flash memory than I know. Which is to be expected–following storage on an everyday basis is his job. I remembered reading about HP’s proposed replacement for flash memory, so I attacked the sky-is-falling-for-SSDs article from that perspective, and he demonstrated that the paper’s argument was even weaker than what I was saying.
And there’s nothing wrong with that. I don’t really care what memory technology my SSDs are based on in 5, 15, or 35 years. I just want them to be faster, cheaper, and more reliable than they were a generation earlier and follow the same path CPUs and DRAM have followed during my lifetime. Hard drives today better than the first drive I bought 20 years ago–at least they’re faster and cheaper, though whether their reliability has improved much is questionable–but they’ve proven they can’t keep pace with the rest of the system. SSDs have a shorter track record, but they’re certainly getting faster, cheaper, and more reliable all the time.
It is entirely possible that the distinction between primary and secondary storage will blur in coming years. That’s more likely to cause the death of SSDs than anything else, though it’s hard to call that death, per se. That’s more like continued integration–it’s like saying IDE controllers are dead. No they aren’t–the modern iteration of them is integrated into the motherboard chipset now, rather than being a separate plug-in card like it was in the days when PC/ATs ruled the world.