Home / News / Peripheral News / USB 3.0 Spec Finalised

USB 3.0 Spec Finalised


USB 3.0 Spec Finalised

While the USB 3.0 specification (a.k.a. "SuperSpeed") was already detailed back in August, it wasn't officially finalised. Now, however, it is and that means we can rest easy knowing that our dreams of 4.8Gb/s transfers will soon be realised.

Okay, soon is a bit of an exaggeration as we're unlikely to see devices using the standard until late 2009 or early 2010. Given the performance figures the wait should be worth it. Real-world testing has a 25GB HD video file transfer in some 70 seconds - a sustained speed of 2.8Gb/s. That's what I call impressive!

USB 3.0 is also unsurprisingly backwards compatible with previous standards, although quite how anyone could cope with the drop from gigabit to megabyte per second transfer rates is beyond me. USB 3.0 also allows larger power draws as well as smaller tweaks, such as a more advanced device-to-device communication protocol, too. In a word, it's just plain better.

It's not all good news though. The long delay in getting the spec ratified and out the door has led to some reluctance in support from many would-be-adopters. Microsoft, for example, has apparently decided not to support USB 3.0 in Windows 7 - out of the box, at least.

USB press release (PDF).

Hans Gruber

November 18, 2008, 10:24 pm

Bits and bytes, buddy. Hugo would not have made such a fundamental miscalculation. ;) The 2.8 gigabit real world speed still translates as some 350 megabytes actual transfer, which is blazingly fast. So fast in fact, it leaves me wondering whether there are any storage devices fast enough to keep up.

So this is what you've been working on whilst replying to your Android vs iPhone write up comments section? I'll let you off then, seeing as how you've been busy and it's almost time for home/pub/off-license or where ever you hard working journos get your inspiration from. :P

Hans Gruber

November 18, 2008, 10:26 pm

Oops! Hugo wrote the story. That changes things (he says, sheepishly). It was only when I'd submitted my off-kilter comment that I realised my fundamental error. Why I didn't check first? :/


November 18, 2008, 11:33 pm

... and the icing on the cake is those speeds can be enabled on a bog standard usb 2.0 motherboard through a simple bios update...?



November 19, 2008, 1:56 am

@ ilovethemonkeyhead:

No, because USB 3.0 has extra conductors (five last time I checked) to enable that speed.

Also, not supporting this in Windows v7 shows how determined Microsoft are to not let *anything* go wrong (and possibly to curtail bloat).


November 19, 2008, 3:39 pm

Azro: My calculation is simply 25 gigabytes x 8 = 200 gigabits &#247 70 seconds = 2.8 gigabits per second transfer speed.

Where on earth did you pull 350MB from? I don't understand your point.

Ryan: More a case of not adding features *most* customers won't care about but might cause problems if you ask me. I doubt the average PC user actually knows about USB 3.0 and as such won't miss its omission but would probably notice if (for some weird reason) the use of USB 3.0 'caused problems with older devices.


November 19, 2008, 4:51 pm

8 bits in a byte.

So ((2.8x1000)x1000)x1000 = 2800000000 bits per second.

2800000000/8 = 350000000 bytes per second.

(350000000/1024)/1024 = 333.79 megabytes per second.


November 19, 2008, 4:52 pm

Azro has a point about transfer speeds. quote: "although quite how anyone could cope with the drop from gigabit to megabyte per second transfer rates is beyond me.". Mixing pear and apples in there; "gigabit" cannot be compare with "megabyte", after all usb 3 at its best 4.8Gb/s=600MB is a "megabyte" connection speed same a usb 2 480Mb/s=60MB/s.


November 19, 2008, 5:23 pm

Haha, just read Hugo's comment... So ignore my last comment.


November 19, 2008, 10:44 pm

@Azro - think that was all me so Hugo has no excuse for, err... still being right!

Hans Gruber

November 20, 2008, 3:23 am

Hugo, I was not challenging the accuracy of your gigabit calculation rate (which you've fully explained above), merely the apparent oxymoron (for want of a better word since I can't think of anything else, I'm not a literary genius like Gordon) in the article's statement (quoted) that people not using USB 3 would have mere megabytes per second data transfer rates than gigabits. By suggesting they wouldn't cope with the drop of speed, you appeared to be implying that somehow data transfer rates at gigabits were better than megabytes, which is obviously absurd (yes you were referring to users of USB2 but my point is exactly the same). 8 binary bits equals a byte. I know you know this but perhaps your statement was misleading. This was my original point. I didn't think it needed further (painful) dissection.

Both Stephen Allred and adulaisow who clearly did correctly comprehend my initial point have explained my own calculation and what I meant (as I've reiterated above). It's nice to have the support of the (smarter) readership. ;) I thought yet again I'd strayed into idiot territory there. Anyway, no need to be so terse. I'm being friendly in case it wasn't obvious.

@Gordon - read it back properly before you comment. Yes, Hugo is right but his statement is certainly misleading or downright wrong (depending on whether he'd confused bits and bytes which I've explained was very unlikely. It still reads that way though). Anyway, after reading this site quite a bit I got the impression Hugo was the more technical of the newscasters between you and he, so made a total idiot of myself by assuming it was Gordon the non-techy who wrote the piece and *apparently* slipped up. So, I'm still right. Hugo's sort of right and Gordon, well mate, you're um, er, hiding behind Hugo? :P

comments powered by Disqus