I swore I'd never use another automobile analogy to describe storage equipment or computing gear after I committed the unforgivable faux pas of referring to a 1957 Chevy Impala in a previous column. More than a few readers were quick to correct me -- Chevy introduced the Impala in 1958 -- and I worried that I might have lost credibility over an old car and one lousy year.
But comparing cars to computers has been a staple of high-tech journalism, and it's really hard to resist.
Solid-state storage has been getting a lot of attention lately, not just on these pages, but in scores of storage vendor announcements. I've said before that solid state is the future of storage, and I'm not backing down from that prediction. But right now, shopping for solid state is akin to fogging up the showroom window while ogling that Ferrari 599 GTB Fiorano with its 620 horsepower engine. For a few moments at least, it's easy to forget that the fire engine-red piece of art costs more than $300,000 and gulps fuel at a rate of approximately 10 mpg. While your heart says "Ferrari," your head knows that the $18,000 Mini Cooper -- with less than a quarter of the horsepower but nearly three times the mileage -- is the car of your future.
And so it goes with solid-state disks (SSDs) right now. Sure, some companies with high-end, ultra-critical apps can justify shelling out the big bucks to buy the Ferraris of storage, but for 99.9% of companies and their applications toiling away out there, SSDs are still a future thing.
One argument says the future is now, as nearly all of the major storage vendors offer some sort of SSD options in their arrays. EMC was the first of the biggies to go solid state and others soon followed, with Hitachi announcing last month that it will offer STEC solid-state storage in its USP V and USP VM enterprise arrays.
There's nothing wrong with those vendors offering high-speed, low-energy storage options for their well-heeled customers but, at best, these are mostly stop-gap solutions -- a way for vendors to say "We're ahead on SSDs," rather than having significant technology advancements. That's not to say customers of these SSD-adorned arrays won't get a bang for their buck, it's just that these arrays represent short-term maneuvers.
A lot has to happen for SSDs to move into mainstream storage. For starters, the price has to come down to where it at least sits on the same budget sector with hard disk alternatives. But the technology itself needs to grow up, too. Although some of today's products make it look like SSDs are a plug 'n play deal, it might not be the best approach. Disk controller systems and software built to run hard disks are likely to need significant alterations to accommodate solid-state storage most effectively.
This isn't to say that storage and solid-state vendors are simply buffing the fenders of the Ferrari. Companies like Fusion-io are hard at work under the hood of solid-state storage, tinkering with the underlying engineering and coming up with more effective implementations. Micron Technology teamed with Sun to address the lifespan issue of SSDs, and recently announced that their joint effort achieved SLC NAND flash that can handle a million write cycles.
Hewlett-Packard is also bullish on solid state, predicting in a recent email announcement that the technology will gain momentum this year. This echoes IDC's prediction of 70% compound growth for SSDs by 2012. Pliant Technology, another solid-state engineering company working on technology specifically targeted at enterprise storage needs, recently announced its participation in SNIA's new Solid State Storage Initiative (SSSI). SSSI is a veritable "Who's Who" of memory and storage vendors, and while the SNIA stamp may not have all that much practical impact at this time, it does reflect its members' level of interest.
You may not see an $18,000 Ferrari anytime soon, but it's probably a good idea to keep your eye on developments in solid-state storage as it travels in the right direction.
Rich Castagna is Editorial Director of the Storage Media Group.