For anyone that may find this interesting, I conducted a test with an external enclosure to see which is faster - USB2.0 or Firewire. The enclosure uses a Prolific PL 3507 chipset for both USB and Firewire so I thought this would be a fair assessment. I used the same 3.2GB file for transfer each time - the only thing that changed was the cable used. The results were as follows (in minuteseconds):
using ordinary USB2.0 cable = 2:07
using Monster USB2.0 cable = 2:00
using ordinary Firewire cable = 1:43
using Monster Firewire cable = 1:38
The results say Firewire is significantly faster. And to those that say those expensive gold-plated Monster cables don't make any difference - well, they did - if only a minor one.
+ Reply to Thread
Results 1 to 8 of 8
-
-
that's very intersting, if you have time could you do an additional test? could you copy a similar amount of data in small files, ~1MB or less? i'd be interested to see what happens there, i get the feeling USB would win out.
-
I'd be interested in a farther-reaching test. Try some different chipsets utilizing the same hard drive. Add some other devices to the chain to see which can handle more IDs. I have a feeling FireWire will come out on top for all of them. I'm pretty close to getting a couple of FireWire800 RAID enclosures. My PCI-X bus will be able to put those to pretty good use and no need to clutter the inside of my computer. The silly thing is that the 1394b SCSI RAID enclosures are actually cheaper than the SCSI RAID enclosures with the 68-pin cable going to them! Now there would be an interesting test
-
Tom's hardware (www.tomshardware.com) has performed extensive testing on Firewire vs. USB 2.0 issue.
Results were no surprise: although USB 2.0 is faster according to standard, in practice, FW400 is faster and FW800 is even better.
Difference is easy to understand, since FW has better coding, less overhead and chipsets are generally more advanced. Depending on used devices, FW-units can "contact" each other directly, USB-devices have to allways use host (usually PC).
In few cases, USB was faster, but difference is likely to be caused by used drive mechanism, since it was accessing little files, when high bandwidth is not such important. Larger files and FW is winner, clearly. It leaves also more cpu cycles for capturing video etc. -
yeah i figured firewire would be faster. what blows my mind is that the better cable was actually faster. i know they are made a little better, but i didnt think it would make a speed difference
-
Interesting... But was anything else plugged into the other USB ports at the same time. I've read that even if a printer is off or a card reader does not have a card in it resources are being allocated to them... and thus reducing the total xfer speed of the active device.
Now, I've also read that USB maintains error checking on its communications where Firewire does not.. Anyone familiar with this? -
Firewire has two different transfer modes, you can read more at
http://www.embedded.com/1999/9906/9906feat2.htm,
it explains Firewire in easily understandable manner. It also has good links for other sites.
One of FW transfer modes has EC, other one doesn't. Bridge chip maker can of course choose, which mode to implement, but it shouldn't be too hard to guess, which one is used. Hard disk transfers uses EC for sure
Therefore, error correction doesn't explain difference in transfer speeds. Fact is that since USB 2.0 is just something "we designed this concept with father-in-law during last week end", it uses all kind of bit-stuffing and because of 1.1 compatiblity issue, it has to transfer all kind of "dummy-data" to have proper size packets.
If there is printer connected, but not turned on, it can't use resources, since "it doesn't exist" to bus. And if you have dedicated port for drive, it's only device in chain, since modern mobos have every connector as separate bus. Bus powered devices, of course, tend to have power on when connected, haven't seen card reader with power switch.
The bottom line: FW is better designed, since it was intended for high speed devices. It has also better wire protocol. That's why it's faster and it shouldn't be surprise. USB 2.0 is just something glued together over USB 1.1. -
1394 trade association is one place worth lookin',
http://www.1394ta.org
If you take a look at FAQ, you understand, why better cable improves transfer rates. It's about signal distortion. Same thing likely to happen with USB.
Similar Threads
-
My blu-ray experiment not going well
By adamf154 in forum Blu-ray RippingReplies: 7Last Post: 10th Apr 2010, 06:11 -
will a USB2.0 Firewire Adapter work for me?
By J.Frost in forum Newbie / General discussionsReplies: 22Last Post: 21st Dec 2009, 16:20 -
Xbox360 subtitle experiment
By BiggerJim in forum SubtitleReplies: 6Last Post: 17th Aug 2008, 08:49 -
Firewire or usb2.0 for external drive media center recording? esata?
By yoda313 in forum Media Center PC / MediaCentersReplies: 29Last Post: 8th Aug 2008, 19:42 -
Best way to experiment
By demonwarrior in forum Newbie / General discussionsReplies: 16Last Post: 14th Dec 2007, 03:11