back to article Intel throws Thunderbolt

Now we know why Intel was dragging its feet with USB 3.0 support - it's got its own competing interconnect that's twice as fast in Thunderbolt (formerly known as Light Peak). Thunderbolt is a 10Gbit/s, bi-directional, dual channel, copper link containing an Intel controller chip and supporting two protocols: DisplayPort and …


This topic is closed for new posts.
  1. Etherealmind

    Tape Drive interface.

    Show be an excellent interface for tape drives in addition to external HDD/SSD etc for better tape drive hroughput and lower cost.

  2. famousringo

    Seems a little negative...

    Not wanting to defend Intel's past anti-competitive behavior, but can you really begrudge the idea of Intel earning billions of dollars on cool tech like this?

    I say we pay the man and enjoy our blistering new data transfer speeds!

    1. Rob Moir
      Thumb Up

      I have to agree with this

      Sometimes an idea is good enough that it makes every sense to start using it. And why shouldn't Intel make money from it? After all they took the risk on investing in the idea in the first place.

  3. HMB

    Complementary Dear Watson


    Why not? In the short to medium term everyone still has stackloads of USB devices that require support. Anyone who thinks USB will go away in a day and a night is sorely mistaken. I'd bet that in 7 years time that even if Thunderbolt is ubiquitous, USB will still be mandatory to support all those keyboards, mice and "USB sticks".

    Eventually of course it will take over, but why ever not? I'm not sure how much I care about the politics when the technology is that good. It's a very clever design too. Expansion cards that you can plug in over PCIe? That's very neat. In fact, I look forward to the day when my monitor, mouse, keyboard and anything else all runs off the same cable.

    Wait a minute, it is a sleek, slim and nice connector right? :P

    1. Anonymous Coward
      Anonymous Coward

      Bluetooth is a great idea too

      ...but how many wireless accessories use their own wireless adaptor instead of bluetooth?

      ...and why is their average price - from what I've seen at east - about 1/4th that of their Bluetooth equivalent (think wireless keyboards, mice, etc)?

      Assuming this great new tech realy is as cool as it sounds (and it does sound very cool) I doubt we'll see any better adoption than we saw with firewire due to cost. I could be wrong of course, but that's what my gut reaction is... FWIW (i.e. not much :)

    2. ThomH Silver badge

      They could just turn it into a breakout box

      USB PCIe cards cost, what, about £20? So — subject to licensing fees as yet unclear — you're basically talking about being able to do a USB hub that connects via Thunderbolt rather than USB for close enough to a normal USB hub that those who want one can buy.

      I guess it'd be the same position Mac people who bought the original iMac found themselves in with respect to their ADB, serial and SCSI interface devices. Except without the forced hand.

    3. Bowrolio

      Thunderbolt and USB support

      I believe you can connect USB devices to thunderbolt connectors with an adaptor - effectively eliminating the need for USB ports of anything for thunderbolt.

  4. Dick Emery
    Thumb Up

    Powered from port makes this a winner

    I was never happy with the fact that with both eSATA and USB2/3 (and indeed firewire) you always had to power something like a 3.5" hard disk from an external PSU. If this can boost speeds and supply enough power to peripherals without having to have another damned PSU in your powerstrip then to me it wins hands down. However now we will have to wait another upgrade cycle to see if become commonplace since USB3 is already doing the rounds and is unlikely to be killed off any time soon.

    1. Michael Habel Silver badge

      Um excuese me sir but,

      You might have missed the Memo where IEEE1394A (e.g. FW400), was actually powered

      Taken from Wiki:

      FireWire can connect up to 63 peripherals in a tree chain topology (as opposed to Parallel SCSI's electrical bus topology). It allows peer-to-peer device communication — such as communication between a scanner and a printer — to take place without using system memory or the CPU. FireWire also supports multiple hosts per bus. It is designed to support plug and play but not hot swapping. The copper cable it uses (1394's most common implementation) can be up to 4.5 metres (15 ft) long and is more flexible than most Parallel SCSI cables. In its six-circuit or nine-circuit variations, it can supply up to 45 watts of power per port at up to 30 volts, allowing moderate-consumption devices to operate without a separate power supply.

      So Pic. related

  5. 142


    displayPort wasn't designed for this, was it?

    ... is this going to be another case, like firewire, of a poorly thought-out connector carrying shedloads of current? I ran a music production college with hundreds of devices connected via firewire to Macs.

    We lost over 50% of the firewire ports on the Macs, and the devices.

    If you disconnected a 6-pin connector with any angular pressure at all, you would short the data terminals on the cable, to the power terminals on the Mac...

    please tell me intel have put some thought into this connector choice...

  6. Anonymous Coward
    Anonymous Coward

    bets on the first Thundernolt peripheral

    One can only wonder how long until we get a USB3 hub that will plug into these. Though USB2 hubs that plug into these would certainly go down well on many levels. Given it would be another optional extra laptop owners can sell.

    Still all this great technology and portage still has to depend upon wall plug socket 1.0 which has been around for years and years :).

    Don't supose they added optional control at the hardware level that enables you to only allow certain devices to be plugged in and work. You know as in a office/work enviroment were you dont want people plugging in there own USB sticks and only company USB devices. Probably not given the culture to DRM protect software and not give any thought into the hardware security for the users perspective. Wonder if you can do a PCIe bus logger with this port, hmmm.

    Still only good can come from this and with Apple biting we will certainly have the peripheral support form the 3rd parties crop up. External HD and I can just smell a USB2/3 adapter being among the first.

    But a poor marketing name "Thunderbolt" that realy does upon it up for some good lines if there are `issues`.

  7. Daniel von Asmuth

    IP over Thunderbolt?

    How does the performance stack up against IP over 10000baseT of yore?

  8. Anonymous Coward


    Well ain't that peachy? Another interface faster than your typical external hard drive.

    It's getting to the point that swapping vids with the chaps at work takes all day, and then some.

    1. Anonymous Coward

      ^^^^ THIS

      Was thinking the same thing.

  9. Kevin McMurtrie Silver badge

    You do not hold cable. Cable holds you.

    It's disappointing that this seems to have become another fat cable technology rather than the sleek optical cable it started as. The DisplayPort cable on Macs is more rigid than Cat 6 and USB 3. It won't stay put and the tiny jack strains to support it.

  10. SAP Bod

    Well done intel + USB isn't going anywhere

    If you think this will oust USB ports you've got another thing coming. Just look around the back of a desktop PC and you may see a parallel printer port, PS/2 keyboard and mouse sockets, VGA socket. These have all been around for rather a long time and all have successor ports (2 of which are USB ironically). The sheer volume of USB devices available means that any manufacturer that EXCLUDED a USB socket would instantly rule out connectivity to a lot of things people already own - so it isn't going to happen.

    To be honest, the writer is painting intel in a very black light, almost to the verge of suggesting how DARE intel release a new technology, the greedy monopolists! Don't be so stupid, stagnation benefits no-one.

    The only question now is whether the market chooses this over USB3.0; but they seem to not even be in the same ballpark and there's no-reason that manufacturers can't implement both. Personally, when USB 3 came along, my only reaction was apathy; woo-hoo faster storage. This has actually got me wondering about all sorts of use-cases and features one of the best bits of FireWire over USB; daisy-chaining!

    Anyone who invents new technology which then succeeds, deserves to be rewarded (to the victor go the spoils). But even if Thuderbolt / Light Peak falls on its arse, I applaud Intel for trying - it's stuff like this that makes the technology sector interesting to be a part of.

    1. streaky
      Paris Hilton


      Yeah and the storage isn't actually faster is the worst part.

      Why it won't kill USB? Well asside from I'm pretty sure USB3 has an optical option anyways going forwards (don't quote me on that) USB4 followed by 5 et al will be around before anybody cares - and it'll probably always be backwards compat AND they were first to market at faster-than-anybody-cares-about speeds. I know this because I have USB3 on my laptop already and no sign of this.

  11. Anonymous Coward
    Anonymous Coward

    Sounds good to me...

    Lets face it, there's still PCI slots on motherboards sold today, 8 years after PCIe was released, to support legacy hardware.

    The same will apply to USB if interconnect becomes a standard.

  12. ph0b0s

    We can forget eSATA as well.

    The same was said about USB 3.0. Anything that is not SATA is going to be slower, as unless something changes and HDD's change to another interface they are always SATA. So you are only ever going to get the speed the SATA / eSATA interface can deliver. Plus you get the extra latency of the protocol conversion from SATA to USB 3, Thunderbolt etc. It may not be much, but it is still there.

    You can argue about eSATA not being a very practical connection, with lack of support etc. But for speed there is nothing faster.

    As far as I am aware.....

    Obviously for non storage devices this will be different, but in those cases eSATA, would not be an option anyway.

  13. Ammaross Danan


    "Intel says a high-definition movie can be transferred across it in less than 30 seconds (neglecting to tell us the size of the file)."

    A quick napkin-math calc shows: 10Gb/s = 1.25GB/s. We'll chop off the 0.25 for overhead and other mythical roundings (basically assuming only 75% is useful throughput), leaving us with ~0.94GB/s. 30 sec transfer at ~0.94GB/s comes out to 28.2GB of data, so they likely considered it a full 30GB Blu-ray rip.

    Having the capability to push PCIe out of the computer, in addition to a USB-type serial connection, will be a very nice addition. However, I doubt a laptop with only 1 of the connections will be useful. Quite likely USB will still have ports, considering using a laptop cooler, mouse, and USB stick eats 3 ports in themselves. I personally wouldn't want to carry around a LightPeak hub just to use more than 1 external device. I'm just hoping they allow AMD motherboard makers to license the tech.

    1. Rob Moir

      not sure you need a "Lightpeak hub"

      - one of the big selling points is that you can daisy-chain devices together. So in your example, all the devices you mention could run out of one light peak port.

  14. This post has been deleted by its author

  15. Anonymous Coward
    Anonymous Coward


    My experience with mini display port on a mac mini is that it doesn't stay connected if moved much. Would some latches have been too much to ask for on an everything connector?

  16. E 2

    vs. other interconnects

    " transfer protocol into a host's memory or PCIe-connected peripherals..."

    I have not read the Light Peak spec. That said, if LP transfers packet oriented data with full management channel(s) in the sense that Infiniband does, then LP might be a decent alternative to IB, modulo existence of a software stack (something like OpenIB), to which an MPI stack could speak.

    If these things were true, this would be very nice because LP seems t o be targetted at the mass market, thus ought to be a lot cheaper than IB hardware.

    Thoughts, anyone?

This topic is closed for new posts.

Biting the hand that feeds IT © 1998–2021