To Thunderbolt or not to Thunderbolt, that is the question—at least for system builders pondering peripheral interfaces. The answer today may be a partial, and qualified, yes. A stronger affirmative may lie in the future, assuming that Intel, the Santa Clara, Calif.-based chip manufacturer, makes good on its backing of Thunderbolt, its new and proprietary peripheral interconnect technology. One key will be the cost.
“The rule of thumb, several PC OEMs have told me over time, is that any new technology that costs over $10 is not going to move toward a 100 percent attach rate,” says Shane Rau, research director for computing, networking, and storage semiconductors at analyst firm IDC. That low cost has to be accompanied by a near-universal desire for the technology, he adds. An interface will then become mainstream and widespread, much as USB has done.
Currently, though, Thunderbolt is not cheap enough or in demand enough for the mass market. So expect to see it in high-end consumer products. In fact, Thunderbolt made its commercial debut this past February in Apple’s updated MacBook Pro lineup. The technology enables 10Gbps data rates, with high-speed data transfer (PCI Express) and high-definition video connections (DisplayPort) on a single cable.
“The big driving factor that will catch a lot of people’s attention is the speed of external storage devices,” says Jeff Hasenauer, vice president of engineering at VAR and system integrator Intellicom Computer Consulting Inc., in Kearney, Neb. Intellicom personnel could use Thunderbolt to set up a server more quickly or work with monitors more easily, notes Hasenauer, who admits that the same may be possible with USB 3.0. Plus, any new interface requires add-on cards for older systems.
Of course, having Thunderbolt on a computer does little good without a device on the other end of the line. And IDC’s Liz Conner, a senior research analyst, storage systems and personal storage, says that products with the new interface didn’t start shipping until earlier this summer. The intended markets for the technology are companies with high-end computing needs, such as design studios and small businesses that work with large volumes of high-content data. “They want to be able to port these huge files from their storage instantly instead of having to wait,” Conner notes.
In the short term, the majority of users—and therefore vendors and their products—are likely to migrate to USB 3.0 rather than Thunderbolt, Conner predicts. USB 3.0 has a head start and offers more than enough horsepower for many consumers, and both Intel and AMD are integrating it into their chip sets, meaning it will be next-to-free to implement.
Longer term, Thunderbolt will need to transition to the mass market. Otherwise, it may suffer the fate of the dying FireWire, another interface that started out on some Mac products but was overtaken by USB 2.0.
The payoff of mass-market success for Thunderbolt could be substantial, since it replaces several different ports and protocols. It could become a universal wire, allowing a future USB 4.0 or SATA 6 and beyond to run over the same cabling. Achieving that much bandwidth may require going back to the future, moving from the current copper-based incarnation to Thunderbolt’s originally envisioned optical transport.
The consensus for now is that Thunderbolt will ramp up slowly. Conner believes that most storage vendors will announce only one to two products with the new interface over the next year or so, doing so mostly to fill out a product line. The products on which Thunderbolt does appear will be high-end affairs, aimed at prosumer applications.
So for system builders and integrators, deciding when and how to leverage Thunderbolt comes down to answering fundamental questions, says Brian O’Rourke, principal analyst at technology analyst firm In-Stat: “If you’re a PC OEM, what you’re going to say is, ‘Do I want a Cadillac or a Chevy? Do I want to pay for Thunderbolt or are USB 3.0 and DisplayPort 1 good enough?’”