Thursday 6 March 2008, 4:32 PM
Why SSD makers don't talk SSDs up
There is a reason for this mismatch, and this is the dirty secret that SSD makers don't like to talk about, at least not in public: namely, that SSDs are simply still too unstable and too slow for the mass market, and the manufacturers are still struggling to figure out which technology to go with.
There are, I am told by the manufacturers, two options: the safe way with single-level cell and the hard way with multi-level cell technologies. Single-level cell is simpler, faster, and has higher endurance but costs a lot more to manufacture. Multi-level cell SSDs are the opposite: mroe complex, slower and cheaper.
Finding the right solution is crucial: SSDs hold the promise of reliable storage (though as we demonstrated in Dialogue Box last year, they do have a peculiar reaction when exposed to radioactivity. They hold the promise of lower power consumption and higher I/O density than hard disks, which is as important for servers as for laptops.
If the manufacturers get it right there is a lot in it for them too; SSDs will command a significant price premium for some time, though once the manufacturing problems are sorted out this could well decrease but even if it does so there will be margin in the business again, which is something that hard disk markers haven't seen for years.
Update (06/03/07): One extra thought about SSDs, again, from talks with SSD makers who would really rather not be quoted, is just how important this could be for Vista. Vista's performance (or lack thereof) is well documented, one reason being that the operating system spends a lot of time grabbing bits of data from the hard disk, even with a decent-sized swap disk. We have cheap, powerful processors and memory, front side bus speeds getting faster, fast graphics cards with features such as SLI – and then we have the hard disk. It's essentially the same technology we've had for 50 years, and the last moving part in the computer - aside from the fan.
I've heard from more than one person in the SSD business that their products will, when they get the technology right, make Vista fly. I'm inclined to believe them because they are not yet, as previously noted, pushing SSDs hard into the market. It's all about getting the mass production right.
Comments on this post
Hello,
the link to "exposed to radioactivity" does not work. How can I get the content of this article?
Thanks,
Wieslaw
My mistake, the URL to the radiation experiment on SSDs should work now. Here it is again just in case:
http://resources.zdnet.co.uk/articles/video/0,1000002009,39287414,00.htm
To add a comment, fill out the form below


