Computerworld is predicting that the end of the line for SSDs will be the year 2024.
That’s based on the projected year MLC flash memory becomes impractical to continue producing. There’s one problem with that assumption: it assumes SSDs will still be based on flash memory in 2024.
Because if HP has its way, a new form of memory it invented called Memristor will go into production sometime in 2013.
Some people, for reasons I fail to understand, want SSDs to disappear. Here’s a news flash for you: Flash memory is the reason your MP3 players and digital cameras keep getting smaller, yet hold more data than their counterparts from a decade ago that used hard drives. It’s not evil.
It’s also not perfect. But it can be made to work well enough today, and its problems are no great secret, so there are companies working on long-term replacements for it.
Memristor may come in and save the day. Or it could be a nonstarter. But 12 years is a very long time. Twelve years ago, 1 GHz CPUs were something new and only bleeding-edge people ran fully 32-bit operating systems. That was the state of the art in desktop computing. Today, we have a name for systems like that. We call them entry-level smartphones. We get them for free (or for a pittance) if we sign a 2-year contract and we can carry them everywhere we go.
By the end of the year, I intend to have 100GB or larger SSDs in all of my systems. The difference in performance they make is so great I can’t stand using a system without one. You should also know that after two months of everyday use, my Kingston SSDnow V+100 still believes it has 100% of its life expectancy left. SSDLife predicts it will turn into a pumpkin in May 2020. I’m not sure how useful it will still be in May 2020. And the V+100 is considered a mediocre drive.
We have 12 years to solve the problem of drives like my V+100. Given what I’ve seen in my lifetime, there’s a chance we’ll solve that problem in more like 12 months. I wouldn’t bet on it, but I’d be even less willing to bet against it.