Storage

For Apple, the future is Thunderbolt

Erik Eckel explains why he thinks USB connectivity is on its way out and how Apple is betting on Thunderbolt technology to take its place.

USB isn't dead. Just recently I reviewed improvements Apple's made to new laptops upgrading USB ports from the 2.0 to 3.0 standard. But the handwriting is on the wall: USB's days may be limited.

What's taking USB's place? Thunderbolt. And before we start a discussion flame war, I'm not talking about Thunderbolt replacing USB this year. But Apple's increasing dedication to Thunderbolt (the company's new Retina-display MacBook Pros include multiple Thunderbolt ports and even the smaller 13" MacBook Pro and MacBook Air models boast one) reveals its intentions.

The company is fueling Thunderbolt efforts. In addition to adoption within new MacBook Pros and MacBook Airs, Mac Minis include the connection technology, too. The company's Thunderbolt display leverages the standard's advantages, also. Meanwhile, TechSpot reports Apple Thunderbolt partner Intel is supposedly working aggressively to improve Thunderbolt performance and adoption, where efforts currently are targeted at "enterprise and other commercial applications."

Is USB an antiquated technology?

Certainly, the ubiquitous presence of USB peripherals means USB connections will remain for a period, but USB isn't aging gracefully. Even with USB 3.0's data transfer capacity, it's outright embarrassed by Thunderbolt.

Courtesy of http://www.apple.com/thunderbolt/

Consider the numbers: USB 3.0 can transfer data at 5 Gbps. Thunderbolt boasts 10 Gbps in both directions (with potentially 20 Gbps on the near horizon). And, if we're realistic, most USB peripherals in use are running at just USB 2.0 speeds, which can only transfer information at 480 Mbps.

Why should businesses care?

Data backups, file transfers, video production, audio editing and other common, everyday tasks are consuming increasing amounts of time as organizations increasingly create, edit and back up additional gigabytes if not terabytes of information. It's no longer unusual for my consulting office to have to back up several terabytes of data a day for a client.

Thunderbolt, clearly the bet Apple is making for the future with its abandonment of FireWire ports and even the cannibalization of USB ports on new systems in favor of the new Thunderbolt technology, helps all those tasks complete more quickly. The result? Staff time is freed to concentrate on other tasks and for moving on to fulfilling other responsibilities and obligations.

USB traffic is subject to congestion, too. Daisy chaining USB peripherals isn't something many IT pros recommend or even possess the capacity to realistically implement. Gone are the days a seasoned IT pro would willingly consider deploying a USB hub. The technology just slows unacceptably or introduces other issues (I've seen printers that don't replicate colors properly, disks that don't run at proper speeds and similar issues when USB devices are connected via a hub). With Thunderbolt technology, that issue is largely eliminated. Some six Thunderbolt devices can be safely daisy-chained.

USB's glory days are ending

Three years from now I may re-read this column and declare "what was I thinking?" You may light me up a few years from now, too, saying the same thing.

But I don't think so. USB connectivity has served the computing world well, but Thunderbolt, with the proper backing from computer companies and device manufacturers, is proving a better technology.

One thing's likely a safe bet: you know how you feel now when you connect a device to a USB 1.0 system? There's that dialog box that appears, the one warning everything's going to run slower in order to enable backward compatibility? I'm comfortable saying now that you're going to feel the same way when connecting USB 2.0 peripherals three years from now. Time will tell.

About

Erik Eckel owns and operates two technology companies. As a managing partner with Louisville Geek, he works daily as an IT consultant to assist small businesses in overcoming technology challenges and maximizing IT investments. He is also president o...

26 comments
JonnyDee
JonnyDee

This article is incredibly biased. For a start, Thunderbolt is a proprietary technology, and has presumably been developed in conjunction with Intel in order to make problems with Microsoft as much as offering users anything. The author should remember that there are billions of little bits of wire connecting all sorts of equipment using USB plugs and sockets. This is just another case of an attempt to divide and conquer, as has been seen throughout the history of IT whenever a supplier things that they are strong enough in the market to get away with it. Looking at raw speeds, a factor of two is hardly worth ditching billions of bits of connecting cable. Apple is just continuing at its dirty tricks of making everything proprietary under the guise of making it slick.

jon.catt
jon.catt

"Data backups, file transfers, video production, audio editing and other common, everyday tasks are consuming increasing amounts of time as organizations increasingly create, edit and back up additional gigabytes if not terabytes of information. ItÂ’s no longer unusual for my consulting office to have to back up several terabytes of data a day for a client." This is true, but not over USB. Backups are taken over eSATA, iSCSI, SCSI, or LAN connections to storage systems such as DAS, NAS, SAN, LTO. The size of data you're talking about would be stored and backed-up from a server device with DAS (direct attached storage) or Tier 1 storage devices (quality SANs) - and not from a client laptop. Any business operating in the fashion you suggest is IMHO "doing it wrong". Thunderbolt will not penetrate the general server market.

realvarezm
realvarezm

Everytime somebody uses something "advanced" and means more buck and what not for the ende user. The same was told about bluray being the new standar! Or what about RIMM memory and lastly but not less important solid HDD. And please dont tell me it will happen in the near future...is all a way for the corporations te get some xtra dollars out of your pocket. Apple do that and they so cinical that they use teminology like innovation, edge and vanguard so when an idiot reads about it, these letters will cause so much curiosity that indeed will spend the $80 for the new slaving device.

CVN76
CVN76

And Top $$$$ for new cable's no thank you! CVN76 and the voice is a beta right.

Komplex
Komplex

and getting leapfrogged when the cheaper technology caught up. Apple used to be exclusively SCSI devices, IDE caught up and SCSI lost out. Firewire vs. USB is another. And I'm betting we can come up with a dozen more. For a technology to be replaced the new technology needs to solve the shortcomings of the old technology. USB Mice/Keyboards/Printers/External Hard Drives/MP3 Players are all light years better than the old PS/2, Serial, Parallell and a few more interfaces I'm forgetting devices. For thunderbolt to succeed, somebody will need to come up with a killer consumer application of the technology. USB 2.0 drives are good enough. Mice, keyboards, external DVD's are good enough. We really don't need to replace them. If all you can say about the thunderbolt is that it's faster than USB, there only a limited market for that. It won't be enough to penetrate the consumer market and unseat USB.

vsmith1
vsmith1

Historically FW was a better interconnection - more reliable, able to carry more power, better sustained data transfer speeds; but USB was cheaper to implement. FW was adopted by others (usually for camcorder connections but was also on Wintel-machines). USB2 did start the erosion of the benefits of FW. Futures The use of Thunderbolt - which is developed by Apple and Intel - is the promise of even faster and reliable interconnections. USB3 hasn't got going. The marketplace is not only about the consumer who is often using wireless connections (WiFi, BT, etc.) but also enterprise (who are moving more to VDI, etc.) but there are also the specialists who want interconnections that are fast, reliable, easy to connect. The future is large data - whether video (moving to higher and higher definition); audio (think of recording with higher definition audio with lots of tracks); even things that we did not imagine (off-loading specialist data transformations such as an encoding devices) as well as the usual storage needs that grow and grow - not everything can be put into the cloud - even with a good internet connection - editing video, audio, graphics is just too much data to move over a network connection. This is the main future for ThunderBolt and even USB3 for a while.

ReadThis
ReadThis

HAL 9000 isn't quite correct because I've owned a 500GB USB3.0 drive for at least a year. It can also operate using a single USB2.0 port.

Gisabun
Gisabun

OK. In general [not Mac specific], you won't see Thunderb olt keryboards or mice [or if you do, no advantage in speed]. Thunderbolt wireless adapter? Maybe a bit faster but wireless will always be slower than just about anything out there. Thunderbolt port on a digital camera? Not for a while. THat leaves USB drives and keys, scanners and printers Have you seen a printer or scanner with a port yet?

Vulpinemac
Vulpinemac

You may not remember how Firewire became the standard for file transfer between hard drives and video cameras many years ago. USB, while good for many general uses, lacked the speed to efficiently and reliably transfer files back and forth--especially large files such as those created by digital video cameras. For several years you couldn't even get a camcorder with a USB connection--not until USB2, to be exact. USB has been problematical pretty much from its beginning but at the same time as long as you stayed within its limits it was easy to use and faster than anything before it. In the beginning USB peripherals worked with smaller file sizes and short bursts of data which never ran into the bottlenecks of being simply a high-speed serial bus. Sure, clock rates have increased so USB's apparent speed has increased with it, but it's still a tiny pipe getting fed more and more pressure; something's gotta blow. Unless you're into the wiring and protocol level of Thunderbolt technology, neither you nor I know exactly how Thunderbolt is different. One thing I'm aware of is that both Apple and Intel plan a fibre-optic connection on top of the existing hard-wired copper apparently using at least some level of parallel communications to enhance that speed. By offering multiple one-way communications lines more data packets can pass without colliding. But I'm sure there's more to it than just that since we already know there's a live circuit board at each end of the cable. The basic USB protocol is almost 20 years old--ancient by modern computer standards. Serial barely lasted 15 years, parallel little longer and SCSI, which in its way was even better didn't last that long. Firewire blew USB away for speed and reliability, but it didn't have the ability to daisy chain to the same extent that USB could and USB2 managed to just meet Firewire's speed. But now that we've pushed that limit once again, we're finding that USB simply can't carry the load any more--we're losing data that could be critical to someone down the road. USB has lived a long and productive life. Let it now retire to usher in a faster, more efficient descendant.

HAL 9000
HAL 9000

And see that message it's painfully slow. When Apple first introduced USB many failed to see the benefits and the same thing happened when Apple introduced Firewire. What was the one big difference between Firewire and USB? Other computer suppliers adopted USB and actively supported it where as Firewire was sort of an Orphan there unsupported and with few devices to use. It's not a matter of it Thunderbolt is better it unquestionably is faster it's weather or not companies will support it other than Apple which will be it's Saviour or herald it's demise. You need to remember that even USB wasn't widely accepted by component makers till well after Apple introduced it and if Firewire had of not been dropped in the next year or 2 it would have taken off just like USB did about 4 years after Apple Introduced it in the PC Market. Even then it's not so much a matter of if Computer Makers make systems with Thunderbolt Ports it's if there are devices to plug into these ports that will be the Make or Break of Thunderbolt, and for that to happen the PC Market will have to openly embrace Thunderbolt which we'll have to wait and see if they do. Even today how many USB 3 plug in devices are there available. A few Thumb Drives are there but currently no external drives which is the important thing to go with. Col

mike
mike

Why does this sound like the Firewire announcements from a few years ago? I guess it's just Apples role in life to discover "the next best thing"... I just wish they always got it right.

Waikiki1
Waikiki1

Thunderbolt is more expensive than USB in the view of system as well as the cables & connector. That makes products price increase

Slayer_
Slayer_

USB 3.1, supports unlimited speeds! As fast as electrons can vibrate into each other. Whats with the limitations?

robo_dev
robo_dev

www dot macworld dot com/article/1162509/speedy_thunderbolt_devices_slowly_coming_to_market.html Thunderbolt is, on paper, 20 times the speed of USB 2.0, twice the speed of USB 3.0. It is faster than SATA III (6 gpbs), so in theory a thunderbolt external drive would be able to maintain the same throughput as a directly attached drive. From a timing perspective, we have had USB 2.0 forever, and are just getting to know USB 3.0. So now Apple does the 'I'm better' thing (ahem Firewire) and thus we have Thunderbolt. The nice thing about standards is there are so many of them....

steve_schaub
steve_schaub

Unless you are direct connecting from a laptop to an enterprise SAN via Thunderbolt (something I doubt anyone would seriously consider), how many devices are there available that can actually sustain a 10Gbps data stream? Flash drives, maybe, but most users are using USB to connect to external hard drives, printers, iPhones, etc and 10Gbps is totally wasted on those uses.

DWFields
DWFields

While it is true that major backups may follow the paths you describe, individual machines may use a separate backup device such as an external drive right at the desk and even now not all machines have eSATA ports--much less Thunderbolt or some other advanced protocol. USB2 is still the most common PC connector and the vast majority of portable and desktop external drives still use USB 2--perhaps with an additional FW800 port or eSATA. To make the blatant conclusion that Thunderbolt will not penetrate the general server market assumes that technology will remain stagnant at least until another generation of external connectivity comes around. Thunderbolt IS that next generation which hasn't yet been fully realized with its paralleled laser transmission which would make data transfer faster than any form of wireless even over short range communications. Until we can develop some form of FTL protocols, Thunderbolt or an immediate competitor/descendant is likely to be the fastest and most reliable we're going to see for at least the next 5 years or so. USB is more than 15 years old. USB 2 is roughly 8 years old now. USB 3--still at minimal use after more than 3 years since it became available, hasn't replaced USB 2 leaving the field wide open for a faster, more efficient level of PC communications.

Slayer_
Slayer_

And people thought those would disappear.

screamino
screamino

Your first line hit it on the head. As techies, we often look at what's best, fastest, most secure, most failsafe, etc., but for the average consumer (and even the average corporate IT mass buyer) cost is often #1, then perhaps security or speed (I'd guess for corporate IT security can often be the first criterion, followed by cost vs. speed). When a breakpoint is created where the new technology is barely more expensive than the old technology, but is unquestionably faster, WITHOUT major accessories and/or modifications, that technology starts to take over. The biggest problems Apple had with Firewire were that only its own machines (for the most part) used Firewire, and it cost a lot more than USB. It also required a different cord than the cord people were using for all their other peripherals, and peripheral manufacturers, as well as phone manufacturers, were quickly standardizing their cables to a couple USB versions (mini and micro). Thunderbolt has the advantage of having a much wider audience of installed Apple devices, as well as Apple fans who are willing to "take the dive" on the technology, but it still becomes a cost equation. For the small percentage of people who are early adopters and tech junkies, cost is rarely a significant consideration. But for the average consumer, twice as fast for twice as much doesn't usually signify enough value to make the change.

Gisabun
Gisabun

You just goofed. you said Apple is on the cutting edge and then said Apple used SCSI [while PCs and others used mostly IDEs]. Then you said that Apple switched to IDE because they became faster. That isn't cutting edge. Would be if they started with IDEs from the beginning. Also, USB3 ports were on non-Mac before Macs. My mobo was one of the first to have USB3 ports. The mobo was released about 3 years ago.

Gisabun
Gisabun

Try and find a decent selection of Thunderbolt hardware. I checked my usual online computer store and found 12 items [many not cheap]. On amazon.com a Seagate GoFlex 1 TB Ultra-Portable External Hard Drive for $260. The same non-Thunderbolt is $90.

Vulpinemac
Vulpinemac

Very essential. Digital camera files are getting bigger all the time and you don't always have the time or place to set everything down and wait for the camera to transfer its files to your laptop/tablet or even swap memory cards. Data transfers between portable drives and desktops or even external drives take longer each day. That speed is critical for a person on the go. Too fast? Ask that question again in 2 years--5 years. You're going to wonder how you ever put up with that belly-crawling USB snail.

bboyd
bboyd

And USB3 is more costly than USB2. Lack of competition and narrow market would be the real reason for prices.

steve_schaub
steve_schaub

My point is that the highway (USB 3.0 or Thunderbolt) is capable of Nascar speeds, but all we have to drive on it are Model T's (spinning disks that top out at 800Mbps). You are only ever going to go as fast as your slowest component, which right now is the disk. If you are expecting to see 5000Mbps when you plug in your Thunderbolt external hard drive, you are in for some major disappointment. Check out this link for a full explanation: http://blogs.howstuffworks.com/2011/04/13/is-usb-3-0-really-faster-than-usb-2-0/

Gisabun
Gisabun

The cost of a USB3 portable drive isn't much more than a USB2. Ditto to flash drives.

Vulpinemac
Vulpinemac

Any more, not even most. Yes, I do agree that laptop and desktop computers do, but there's not one modern tablet that does; most "ultrabooks" don't; cameras don't; MP3 players don't; printers don't... well, you get the message. As I comment farther down, I ran into the limitations years ago while editing videos for my clients where USB simply wasn't fast enough or accurate enough. I'll grant USB is faster today, but it's also trying to run NASCAR on a two-lane highway.

bboyd
bboyd

Solid State Disks are cheaper every day. Right now the common drives max out a SATA 3 connection. Plus that bandwidth supports times when multiple devices are daisy chained on the interface. Drive to Drive copy is a reason for excess bandwidth. Ever had your WiFi come to a crawl because 5 or 6 devices are poling it a the same time? Plus how much of the bandwidth will be used up by overhead. Interpacket delay, collisions line length and other factors will take the real world numbers down. If there is plenty of excess then it will never be noticed. The connection will always be as snappy as the devices allow. Thunderbolt's spec is far better optimized for these factors than USB.