A hard-drive standard flops outside the box
SATA is ubiquitous for plugging hard drives into computers. But a version for external use has never caught on. What happened?
- Shankland covered the tech industry for more than 25 years and was a science writer for five years before that. He has deep expertise in microprocessors, digital photography, computer hardware and software, internet standards, web technology, and more.
You'd be hard-pressed to find standards as ubiquitous as SATA, which is used to plug hard drives into computers. But its success inside the computer chassis turns out to have been a bad predictor of its success outside.
Years ago, SATA allies created a variation of the specification called eSATA that would let people attach hard external hard drives to computers. The big advantage over USB: an eSATA drive reads and writes data just as fast as an internal drive.
Despite its branding disaster of a name--eSATA stands for External Serial AT Attachment, and AT stands for nothing in particular--eSATA achieved some measure of success. I for one am glad it exists as a way to give laptops some measure of storage expandability of desktop machines. But overall, it never built critical mass, and I believe new technologies that match its speed and exceed its breadth will consign it to obscurity among mainstream computer users.
The nearest competitive threat is the new USB 3.0 "SuperSpeed," which offers transfer speeds of 5 gigabits per second compared to the 480 megabits per second of the currently prevailing version of the multipurpose Universal Serial Bus technology.
The new USB version is just now assuming the throne after a dangerously long reign by its predecessor. The first hard drives supporting it are on the market, and soon it will become mandatory in PCs.
In comparison, eSATA offers speeds of 3Gb/sec, but the only thing you can attach with it are hard drives and optical drives. As a feature on most people's purchasing checklists, it's optional.
In-Stat analyst Brian O'Rourke concludes that eSATA "will remain a niche technology with very limited growth prospects." The market research firm forecasts that external drives supporting eSATA will drop from 8.5 percent of the market in 2009 to 7 percent in 2010 in a new report on the matter.
Where eSATA falls short
I thought eSATA had promise--I use it daily and think eSATA is smart for higher-end storage devices such Data Robotics' new Drobo--but my personal experience has revealed difficulties and complications.
eSATA's incumbent advantage is that the protocol is no different than regular SATA. All computers need to support it is an external port.
But the world is moving toward laptops, where that one extra port is precious real estate that could be used for another USB port or for making a design just that much smaller or thinner. A clever solution on my own laptop combines a USB and eSATA port into one, but that's more expensive than an ordinary USB port.
A second eSATA problem is inconvenience. On my machines, eSATA drives must be powered up before the computer boots. For a stationary computer or video recorder, that's not so bad, but since USB has accustomed us all to the convenience of plug and play, eSATA's finicky nature is a significant drawback.
In my experience the boot-up issue poses problems for some software, rearing its head in particular when the computer is going to sleep and waking up.
eSATA also got off to a rocky start for me. It's working for me now, but the first eSATA drives I bought were erratic enough that I just connected them with USB or IEEE 1394, aka FireWire.
But there's another contender besides USB, too: Intel's Light Peak optical connections. Intel hopes Light Peak will be even more all-encompassing than USB, embracing video and network connections that USB today can't handle.
Why go optical? Transmitting data over copper wires at ever higher speeds poses challenges when it comes to shielding against electromagnetic interference, and the new-generation USB cables must be shorter than their predecessors cables to accommodate the faster signaling rates. In comparison, the optical connections of Light Peak hold the promise of much faster communications--10Gb/sec to start--especially considering the how much room optical communications has to grow.
But it's not clear yet how expensive Light Peak will be. I worry the cost could be high enough to consign Light Peak to another high-end niche, even if optical links are eventually the future.
After all, the industry is littered with interconnect standards that achieved some success but never made it big: FireWire, Fibre Channel, and InfiniBand.
And now you can add eSATA to the list.