USB 2.0 has been stretched to its limits in both speed and power. While its 480 megabits per second (mbps) of raw bandwidth once seemed like plenty, that was before we routinely transferred multiple gigabytes of video clips, photo archives, and design files.
And while USB 2.0 wasn’t initially designed to be a universal standard for charging mobile devices, that’s exactly what it’s become: In a few years, most every cell phone handset will likely have a USB jack. Unfortunately, USB 2.0 can’t necessarily deliver the juice. USB 2.0 can deliver up to 0.5 watt (passively charging) or up to 2.5 watts (actively communicating with a host driver) to devices plugged into a port. Apple’s proprietary high-power USB ports on some Macs can push out 5.5 watts, but that’s only with Apple gear, like the iPad.
The USB 3.0 revision should solve both the speed and the power problems. The spec calls for 4.8 gbps of raw data transfer—10 times faster than 2.0. (Both standards deliver about half that in actual throughput.) Along with the increased speed comes increased power. The USB 3.0 spec allows just under a watt of power for unconfigured devices, and up to 4.5 watts after a device introduces itself.
At the same time, USB 3.0 will remain compatible with USB 2.0 (but not 1.1). All USB 3.0 devices will have separate 2.0 hardware inside. Plug 2.0 or 3.0 cables into a computer with a USB 3.0 port, and it’ll talk at the right speed to whatever device is on the other end. USB 3.0 is worse than 2.0 in one way: The maximum cable length is reduced from 5 meters to 3 meters.
Despite its technical advantages, USB 3.0 has appeared in only a few peripherals so far; for example, LaCie recently released some USB 3.0 hard drives. Printers, scanners, and cameras don’t need 3.0 speed; hard drives do.
Apple has yet to make a public commitment to USB 3.0. In fact, a report recently circulated that Steve Jobs had issued one of his sporadic e-mail responses to a user, saying that Apple had no plans to adopt USB 3.0. As usual, Apple wouldn’t confirm the veracity of that report.
In any case, USB 3.0 is unlikely to appear before OS X 10.7 (Lion) ships next year—if Apple decides to adopt the technology at all.
(One more note: You might have heard about Wireless USB over the last several years. That technology never took off; you can find the ultrawideband-based standard in only a handful of laptops and other devices. While Wireless USB took a slow route to market, Wi-Fi sped up and became cheap and could cover much greater distances than Wireless USB.)
Even though Apple now includes FireWire 800 in some Macs (all
desktops and the MacBook Pro), I don’t think FireWire has much of a future.
FireWire originally appeared at a time when USB 1.1 was crazy slow and Ethernet was stuck at 100 Mbps. Sony and a few other vendors included FireWire in their products (as i.Link or IEEE 1394). But USB 2.0 still found broader acceptance, in part because peripherals that used it were less expensive and came to market in vastly larger numbers.
So even though FireWire has always offered more-reliable, higher-powered, and higher-speed communication than USB, it’s no longer seen as a real competitor. Two faster FireWire flavors (1600 and 3200) have been available for years, but never widely deployed.
I’d expect FireWire to persist until 2011, but then quietly disappear as USB 3.0 takes over. Tower Mac users will still be able to buy plug-in FireWire cards.
If Apple doesn’t adopt USB 3.0, and it lets FireWire die, how will we connect high-speed peripherals like external hard drives? If you ask storage vendors that question, some of them may go off the record and whisper about something called Light Peak.
Light Peak is a new optical-cable technology that Intel is developing as a replacement for USB, FireWire, and even eSATA. It would allow you to connect devices in a peripheral bus with a bandwidth of 10 gbps. Even more impressive: The same basic technology, with new hardware, could hit 100 gbps.
Intel has demonstrated the technology to members of the press and to certain developers, but it’s still some years away from market. The question won’t just be how quickly Intel can develop it, but also whether or not Apple will be willing to implement it.
Bluetooth has become a common connector for keyboards, mice, headphones, and other devices. Its big advantage for close-range wireless connections—compared to, say, Wi-Fi—is that Bluetooth devices can arrange data transfers with each other without requiring any networking infrastructure.
In its initial form, Bluetooth transferred data at 1 mbps; that wasn’t fast enough to support multiple data streams or high-quality wireless stereo audio. The 2.0+Enhanced Data Rate (EDR) flavor supported 3-mbps traffic speeds; the current lineups of Macs come with 2.1+EDR.
The 3.0+High Speed (HS) standard is currently available in a limited selection of hardware. It boosts file transfers to about 25 mbps, while regular usage preserves power and stays at the 3 mpbs speed. But Apple may not have a compelling reason to include it in Macs in 2011 unless it’s added to new iPhone and iPad models.
Current Apple hardware uses the 802.11n Wi-Fi standard. But two new wireless standards are coming, both of which will support much higher speeds.
Right now, 802.11n allows devices to transmit data in two spectrum bands: 2.4GHz or 5GHz. The maximum capacity for the 5GHz band is 600 mbps, but no gear works that fast yet. Apple’s current AirPort Extreme base station and Time Capsule can deliver up to 450 mbps of raw bandwidth in the 5GHz band; as far as we can tell, Macs support just 300 mbps at 5GHz and 150 mbps in 2.4GHz.
The Institute of Electrical and Electronics Engineers (IEEE) is working on a revision that would boost 802.11n speeds in the 5GHz band to more than 1 gbps (that revision is 802.11ac). The 802.11ac spec is due for completion in 2012, but you can expect to see gear supporting it shipping a year or more before that.
The IEEE is also working on a complementary standard in 60GHz—802.11ad—that will offer speed improvements of several gigabits per second beyond 802.11n. But there will be a catch: 802.11ad will likely work only within a single room. That means 802.11ad could compete with USB 3.0 for connecting hard drives and other devices. It would be particularly handy for computer-to-computer transfers.
Another industry group, the Wireless Gigabit Alliance (WiGig) has been working on its own 60GHz standard; it now appears that the WiGig standard will converge with 802.11ad to form a single specification.
A separate standard for streaming high-definition video in the 60GHz band—WirelessHD—is under development, too. It’s almost certain that this standard will coexist with WiGig, but it’s unclear whether both streaming HD and regular data networking will wind up incorporated into one network flavor.
802.11ac won’t be ready for deployment until 2012. WiGig and 802.11ad are unlikely to hit the market before 2013 in any widespread way.
Glenn Fleishman contributes regularly to The Economist, the Seattle Times, and Ars Technica, and appears regularly in Macworld. His latest book is Five-Star Apps (Peachpit Press, 2010).