Do we need 10Gbs USB 3.0?

Do we need 10Gbs USB 3.0?

Summary: The USB 3.0 Promoter Group announced plans to double USB 3.0 bandwidth to 10Gbs. That would make more sense if anyone was complaining about current USB 3.0 speeds.

TOPICS: Storage

I've been using USB 3.0 and Thunderbolt for the last six months. I'll write in more detail about my experience with USB 3.0, but my key finding is that it is not nearly as robust as Thunderbolt.

Also, there aren't that many devices that can make use of the performance of USB 3.0 today. Its 400 to 500 MB per second bandwidth far exceeds that of any single hard drive, and is fast enough for almost any single SSD.

So what is the purpose of this announcement?

What's the rush?

The announcement is surprisingly light on any detail about why faster USB 3.0 is needed. After all, the original USB 3.0 has been incorporated into Intel chipsets only in the last year. Most devices still use USB 2.0, which is plenty fast enough for Blu-ray burners, scanners, printers, and many thumb drives.

However, the announcement is not about the adoption of a new specification. It is an announcement about the intent to create a new specification that should be finished later this year.

And when might it be seen in products? No word on that, but the v1.0 USB 3.0 spec was released in November 2008 and took almost four years to get wide adoption on PCs.

The only clue to why and why now comes from a quote attributed to a Microsoft manager: "These updates will enable higher data rates and allow combining of disk, high-definition audio/video, and networking traffic on a single cable — all while maintaining compatibility with billions of existing devices"

Sounds like Thunderbolt. Except that Thunderbolt today supports 2 full-duplex 10Gbs channels, which is what enables it to support external RAID arrays or PCIe card cages--among other things--and a concurrent 2560x1440 display.

Even with the update, USB 3.0 will have half the bandwidth of Thunderbolt today.

Head 'em off at the past!

To recap, the USB 3.0 Promoter Group has announced a plan to double the USB 3.0 data rate spec, with no use case, no timetable for product,s and a requirement that users buy new USB 10Gbs computers and hubs. And they still won't have the capability that Thunderbolt has today.

Smells like a pre-emptive strike against Thunderbolt 2.0, which, as I noted last year, is expected this year. Perhaps it will debut on the semi-announced new Mac Pro.

The Storage Bits take

The slowdown in PC sales is up-ending old verities. As desktops and laptops become professional tools rather than lowest-common-denominator game/email/browsing devices, they benefit by having ultra-high-performance options like Thunderbolt.

As tablets become the consumer computing device of choice, desktops and laptops have to migrate upmarket to remain relevant. Even though volumes may drop, average sale prices can rise if they provide functionality that professionals require.

I suspect that the USB Promoter Group realizes that USB 3.0 is not needed on tablets, and must compete with Thunderbolt to have a continuing critical mass of adoption and devices. But Thunderbolt is engineered to reach speeds of 100Gbs by the end of this decade.

There is no way that USB 3.0 can compete with Thunderbolt head-to-head over the long term--and they don't need to. But I commend them for trying.

Comments are welcome, of course. Are you using USB 3.0? How do you like it?

Topic: Storage

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • The last PC I bought...

    ... had two (and only two) hard-to-reach USB3 ports on the back, rather like USB ports in the late 90s and early 2000s. And I have a single 1TB USB3 drive that can use them properly.

    Having said that, I've plugged this drive into several USB2 ports and still copied multiple GB of data across in a reasonable amount of time. My overall experience with USB3 to date can therefore be summarized as "Meh".
  • I've no need for USB3.0 personally

    Even my backup drive backs up at 4am. I'm asleep, I don't care how long it takes to finish.

    My laptop has two USB3.0 and one USB2.0 but all have USB2.0 things plugged in.

    Still, I don't see a faster connection (likely to compete with Thunderbolt) doing any harm.
    • If your drive is small..

      Yes, you do no need USB3 if drives your are backing up are small. I recently did full backup of a PC and it took 12 hours. External disk was connected via USB2. Having USB3 would have made backup run faster.
  • Price?

    "There is no way that USB 3.0 can compete with Thunderbolt head-to-head over the long term"

    What about on price? Looks like Firewire vs. USB2 all over again.

    I think the "Single Cable" thing is way over-rated. If its two cables and save $100 vs. single cable, I'd bet the two cable solution wins easily.
    • My thoughts exactly...

      The backward compatibility of USB 2.0 with USB 1.0/1.1 along with lower prices pretty much buried firewire/1394 except for specialized applications.

      I see the same forces in play with regards to USB 3.0 vs thunderbolt.
      • No because

        USB 3.0 is buggy piece of crap! I've gone through 3 USB 3.0 cards and not one of my 3.0 devices works consistently. It stinks to have a2TB hard drive you can't access 70% of the time! Even my new motherboard with onboard USB 3.0 is not much better. And I don't see anything Thunderbolt available in everyday accessories so I'm forced to use USB 2.0!
  • Great quotes from you

    "Also, there aren't that many devices that can make use of the performance of USB 3.0 today. Its 400 to 500 MB per second bandwidth far exceeds that of any single hard drive and is fast enough for almost any single SSD."

    Got it. There is no reason to get anything faster than USB 3.0 today.

    "Even with the update, USB 3.0 will have half the bandwidth of Thunderbolt today."

    So what if Thunderbolt is faster? You just finished convincing us that there is no reason to get anything faster than USB 3.0.

    You can't make this stuff up.
    • He also says

      "There is no way that USB 3.0 can compete with Thunderbolt head-to-head over the long term - and they don't need to."

      So in essence, he IS saying "So what if Thunderbolt is faster?".
      Michael Kelly
    • Hmmm… Yeah that makes no sense really - what is needed?

      Yes I found a reson for needing faster than current USB 3.0.

      TV is going 4K (sets available now but pricey) - which uncompressed at full frame rates will need just faster than USB 3.0 spec.

      Now OK you can do this with Thunderbolt and I probably would - but it leaves USB 3.0 a bit behind. It can stay that way if you would rather limit the consumer, so lobby away if you wish.

      What is needed or not needed today depends on who you are. If thunderbolt does the job then I guess no reason for USB 3.0 is there!

      Oh and if you think 4K is pushing it and not going to happen, 8K is in it's early stages and will follow on.

      4K cameras for consumer use are also hitting the market. So this is not just a pro video issue. I guess most consumers will work compressed though.

      @ toddbottom3 - you often make stuff up. You just had to invent a reason to think you were not threatened by Apple by missing the point (again)
      • apple?

        "You just had to invent a reason to think you were not threatened by Apple by missing the point (again)"

        Thunderbolt is an Intel technology that has nothing at all to do with apple.

        "Yes I found a reson for needing faster than current USB 3.0."

        You should tell Robin this. He is the only saying that USB shouldn't get any faster.
      • I don't know what your experience with professional video is...

        And I wont try to guess.

        However, I cannot see any pro video editor using usb OR thunderbolt to display the work they are editing. There is no way in hell either cable is outputting uncompressed video regardless of what the video content is. The signals coming out of these cables is going to be designed for max compatibility and to reduce the cost of display devices, thus it will be compressed by the hardware. any display device that is capable of showing 4k in its native uncompressed resolution is going to be extremely expensive due to the cost of the internal components. A real pro is using a DVI DUAL link cable on a graphics monitor capable of rendering a full 48bits per pixel color space, and it will almost always be a 1080p monitor.

        Also... 4k is not realistic. most people, and by most i mean 99%, cannot tell the difference between 1080p and 4k at the standard viewing distance of 10 ft. And for anybody that thinks people will buy larger screen sizes so they can overcome the limited spacial resolution of the human eye apparently does not understand that the average person doesn't even know what that is. On that same note, have you ever used a 2k monitor, much less a 4k? especially in the affordable size range of 20"-26"? The text on the screen is unbearably small. Try setting your gui's font size to size 6 and try to get some work done, as this is a good simulation of what it would be like to work on displays of 2k res or higher. your going to find your self hunched over your keyboard trying to read.

        And if your solution to the readability problem is to increase the dpi or font size of the onscreen text, why not buy a regular old 1920x1080 monitor and leave things as they are. what we gain in onscreen real estate is over shadowed by the fact that we cant actually make it all out in any meaningful or productive way.

        Yes 4k is great for movies, but what you will find is that most pros record in 4k (or 8k) with acceptable levels of compression, and then use the extra spatial resolution to produce extra sharp 1080p video. its an old trick that works 100% of the time. and consumer cameras that record in 4k are a joke considering no normal person can afford a tv or monitor that has native 4k resolution. its all hype designed to sap a little more money out of people who think their perfectly fine 1080p cameras are outdated.

        Using 4k as a defense of usb 3.0s future speeds, or of thunderbolts current speed, is pointless since far too few people will seriously think to use either for high res video, even if capable.
    • The Microsoft guy nailed it. . . .

      The great thing about Thunderbolt is that you can do so much with it. I have 2 large monitors and a RAID array attached on a single Thunderbolt port. I will probably add a video capture box as well.

      Which is what the Microsoft guy said about 10gig USB - only since it will still have half of Thunderbolt's current bandwidth I don't see how it will perform as well. Which isn't to say that USB monitors won't come out, but will people buy them when HDMI is already everywhere?

      R Harris
      • Got it

        So USB monitors suck because there won't be many. Thunderbolt monitors are great because there aren't many.

        Wait, what?

        Regardless, here is why faster USB can be better:

        This is a docking station that uses USB 3.0 but the monitors you plug into it use normal VGA and DVI / HDMI connections. No need to buy a new monitor with this.

        My prediction is that Thunderbolt will suffer the same fate that Firewire did. It is a solution that gives 0.5% of the population the bandwidth they need while being COMPLETE overkill for the rest. And it is expensive. Far more expensive than USB. So 99.5% of people get the privilege of paying for something they don't need.

        How ironic that in today's world where people are being told "don't buy a PC, you don't need all that power, buy a cheap Chromebook" you are here telling us "you should pay more for power you are never going to need".

        Robin, you are on the wrong end of the pendulum with this one.
  • Forward thinking

    Some people, but fortunately not ZDNet bloggers, have to be forward-thinking. They have to anticipate what people might need 3 years from now and plan for it. If you want to prevent people from having a reason to consider using a different technology, you have to stay ahead of the curve and let people know what your plans are for your technology. If this is a spec that won't be done until the end of 2013, it will be 2015 or 2016 before the technology makes it into systems. By then, there might be a need. This is the USB 3.0 promoter group letting people know that they are not going to let USB 3.0 become the bottleneck.
    • Good observation: Needed tomorrow.

      Experience pretty much teaches us that you can't have a processor too fast or a drive too big--the next round of software will eat it all up and then some. I expect that long term you can't have data transfer that is too fast. Remember the blazing speed of the giant capacity 1.44 meg floppy compared to the ones that were floppy?
  • Photographers/Cameramen would benefit

    I know they end up needing to offload a dozen or so MicroSD chips at a time...they have the high capacity ones, the size of most SSDs.
  • Compatible with previous USB

    is the killer feature. This means all the current USB devices keep working with no need to upgrade. Of course Thunderbolt would be great, but I don't see that readily available on any of the motherboards/computers I would buy.
  • enough is enough

    Do we really need more innovation? I think we should be content with what we have and stop trying to improve it! My great grandfather didn't even have USB3.0 and he was happy!
    • Congratulations

      You win a guest spot on zdnet with your backward logic. I look forward to an improvement in the general quality of articles.
      Little Old Man
    • Made my morning

      Very funny

      But was it the USB 2 or the meds (or the nurses) that made him happy?

      Perhaps we will never know.