"In some sense we'd... like to build the last cable you'll ever need."
"640K is more memory than anyone will ever need."
Intel sees its high-speed, long-distance Light Peak optical cabling technology as the next interconnect step after USB 3.0. "We view this as a logical future successor to USB 3.0," IDG News Service quoted Intel senior fellow Kevin Kahn as telling his audience at the Intel Developer Forum currently underway in Beijing. "In some …
Though, I will say that optics do offer a way to make a cabling investment that extends long beyond the life of the termination equipment. Of course, if this becomes the future standard, an investment in Corning is probably warranted too, to cash in on all the fiber optics required...
About 15 years ago, I my monitor resolution was 1600x1200 at 85Hz (4Gb/s). Although the resolution has gone up to 1920x1200, the huge frame rate is no longer needed to prevent flicker so the bandwidth requirement has dropped to 3.3Gb/s. Years from now, when 3D films become common, the first generation of Light Peak will have no problem shifting a pair of uncompressed high definition videos. A single 100Gb/s link could handle a 4x4 grid of high definition 3D monitors. That is so big I could not put one in my living room. Perhaps weather forcasters and nuclear bomb simulators will want more, but I do not expect to need over 100Gb/s at home.
I feel USB3 will meet the majority of my interconnect needs and that like USB2 has become, be inexpensive in doing so. Shrink the connector way down? They're already small enough that one wrong bump can damage a device beyond reasonable repair. IMO if your device can't manage a cubic centimeter of space for a robust connector, if that's too bulky for your tastes then you probably didn't care enough to take it along with you regardless of such a small difference in size.
"""IMO if your device can't manage a cubic centimeter of space for a robust connector, if that's too bulky for your tastes then you probably didn't care enough to take it along with you regardless of such a small difference in size."""
Actually... That's entirely wrong. A connector doesn't just take up space, it forces other components further apart, which tends to cause devices to become rather thick. This is why micro-usb (Which is on all of the current devices... which I guess you would never use. That leaves you stuck using... an iPhone I guess.) isn't narrower than mini-usb, just thinner.
An honestly, definitions of 'Robust' change depending on the application - certainly you don't mean that everyone should have a canon plug rated for use in explosive environments on their phones.
"""They're already small enough that one wrong bump can damage a device beyond reasonable repair."""
Not sure what you define as one wrong bump, but I'm going to go out on a (short, sturdy) limb here and say that you're full of shit on that one. I'm not too delicate with my devices, and I've owned quite a few with USB ports of all sizes, and I've yet to break any, even by regularly picking things up by the USB cable.
I don't see why people have to rally against anything new. Your worry about cost effectiveness is moot, because if it isn't, then manufacturers won't pick it up, and consumers won't spend extra for it, and it'll go the way of the glorious firewire port. That means that if/when we see it available, it will already be cost effective. And technically, if it's the only thing that offers 10gbit, then it'll be cost effective for some people. If you will /never/ need 10gbit, then go ahead and don't use it, simple as that.
I for one can't wait to see USB get relegated back to what it was designed for, and the only thing that it's good for even after all the hackery of later versions - low speed / low power peripherals. Assuming that Intel can work out a slightly less broken protocol for their new optical magic, it could be the first thing to legitimately replace firewire on grounds of actual usefulness.
If this thing is as good as it sounds, and can push HD Video, Network, etc all down one (potentially very long) fiber, then I want it yesterday, and I can't see how anyone else wouldn't. If it lacks the irritating master-slave setup that makes USB so useless, it could make for some very interesting networks, and solve a whole lot of irritating problems.
The beauty of Light Peak is that, if you pardon the expression, it's one cable to rule them all. It also only requires two components - the optical transport and the power feed rather than a cumbersome, fragile, multi-pinned, which-way-is-up?, interference prone connector and limited length cable.
On small/ultra-thin devices you could get away with one tiny port allowing even smaller and thinner kit, and on computers instead of having a plethora of different ports (USB/Firewire/Audio/video/ethernet) you can have just one or more tiny Light Peak ports which, with breakout boxes or adapters, could connect to absolutely everything.
Bring it on!
But one of the great things about usb is the ability to power low draw devices with it. That slimline hard drive's going to be a bit tricky to hook to hook up your laptop on the train. But I guess it could always have a second socket to power it. Hmmm that usb port's not doing anything, let's hook into that. Oh. Erm.
I agree, there should be some sort of copper (or other conductive material) cable to provide power. If it doesn't provide that, some people may not want to switch.
Also, Cat5 cables can transmit power. Kinda how some landline phones get their power over the phone line, some network appliances, like wireless access points, can draw their powers. There are several switches that provide Power over Ethernet.
One other thing I really hope Light Peak offers is a controller chip like Firewire & some ethernet cards have (like a TCP offloading engine). This way, we can have peer to peer networks, and 10-100 Gbps is a lot of data. If there's a lot of overhead, that could really slow transferring data down.
Also, I hope whatever protocol they use for Light Peak is more efficient than the TCP/IP stack. The TCP/IP stack was made back when people were mainly using dialup connections instead of broadband. Now, it's pretty inefficient. And it probably won't change any time soon considering how ingrained it is in the internet. I know, we're going to IPv6, which is better, but it can still be much better.
"you could get away with one tiny port allowing even smaller and thinner kit, and on computers instead of having a plethora of different ports"
Wasn't this supposed to be the raison d'etre of USB ? To get rid of PS/2, serial, parallel, external SCSI, etc ?
I'll grant you that it's partially achieved this, but not totally.
Thjat maybe what USB was*meant* to do, and i dont know about you, but looking back at the good old days of mouse, kbd, video, modem and just the2 speakers and a couple of mains leads makes me smile, now because USB is so convenient (whaddya mean it's only got 12 ports on it??) I can hardly see my beloved box for the millions on damn cables for all the extra kit that is now on there.
the important message to learn from this is " crap will expand to fill the available space"
also at the moment, my phone, camera, both types of keyboard, wacom, mouse memory sticks etc all use usb to provide the requisite nimble amps to make them work as well as a data connection i'm pretty sure that this optical tech will not be able to do that, so more wires then :-)
yup just like the paperless office!
Mine's the one with 30 yards of assorted wires and adaptors hanging out the pocket
Being optical, wont the fibres break down quicker then Copper?
Also I see combo cables being used Copper for power, Optical for Data, also I think it would be cool for the devices at each end to perhaps have IP address so that some Computer systems can daisy chain lots of stuff :)
For example A PC then Flash drive then Router, then printer all using IP address, means anything can connect to anything! Of course this would keep sys admins busy!
Fibre optics are the best to offer consumers. They're so brittle that they won't last five minutes in the average family home, and a lucrative after-sales industry of replacement cables will spring up.
PC World must be salivating already. And they can add an 'eye test' to the sockets on your PC when you book it in for an 'expert' health check, and advise a set of 'special fibre optic glasses' to fit over the end of the socket for old PCs whose fibre optic vision is probably starting to get blurry. All for just £39.99.
"Fibre optics are the best to offer consumers. They're so brittle that they won't last five minutes in the average family home, and a lucrative after-sales industry of replacement cables will spring up."
Since when are consumer aimed fibre optics brittle? I've got a bunch at home for digital optical feeds and they are far more flexible and maliable that any copper based cable I've used.
Bear in mind, conductive cables usually mean copper (often tin coated), bend copper around a bit and it breaks, hence why copper cables are usually made up of many fine strands, it's not because this stops it from breaking (reduces yes, stops no), it's just so it breaks in each strand in a different place so you don't loose conductivity.
Modern optical cables are made of polymers that can't be bent to a very high degree before being damaged, i.e. squash it under foot etc. So should last an awful lot longer than any copper based cable.
Well remembered sir!
The Apricot F1 and also the Apricot Portable (a big black Toblerone with an LCD) had IR keyboards and trackballs. The protocol was very simple and there was a risk that adjacent systems in an office might cross-couple, so fibre optic cables were provided to ensure they didn't.
is Intel trying hard not to let LightPeak die. They tried to convince everyone that LightPeak was the future now, and no one was buying it.
They refused to produce USB 3.0 motherboards and the industry went around them. Now they've realized that they cannot hope to stop USB 3.0, so they are trying to position LightPeak as a successor.
Maybe, maybe not. USB 3.0 is just getting started. Let's not forget that eSata looks rather good right now as well, for storage devices only. I also expect eSata to make the jump to 6g/s SATA-3 when it's needed in 1-3 years.
Fiber Optics are more expensive than copper in every implementation I have experience with. (Toslink, Fiber SC, Fiber LC.) They are more fragile as well, and they definitely cannot carry power. So I see this technology going nowhere in the face of USB 3.0 and continued progress with eSata.
"Since when are consumer aimed fibre optics brittle? I've got a bunch at home for digital optical feeds and they are far more flexible and maliable that any copper based cable I've used."
OK, I DARE you to bend one of those wires RIGHT BACK on itself!" YES I mean a complete U-turn! :D No not the large diameter ones, the typical usb ones that are less than 3mm thick...
the copper cables may stretch and strain, but they will still work.. my one has been thrown about, shut in doors(some permanently, due to a landlord not wanting holes drilled in the door surround!!), etc, etc, but still works..
Of course there will always be copper for the power supply, there already is in USB - If that was not needed, there would be only two wires, like audio..
It'll be interesting to see a future LightPeak 4 port Hub to take all my external devices.
Of course it would have to have backwards compatibility with my then ageing USB 3.0 devices otherwise why would I bother?
That'll make it quite expensive incorporating all that electrical to light signal conversion components.
yes, very true... another thing to think of though, is that most reviews of USB3 say that throughput may well be very fast, BUT even the latest HDD's speed is now the bottleneck... and then there are things like latency of the interface, software controlling it, etc...
All the same things we heard about USB2, yes... The real story about USB2 is that yes, it is fast for many small files, BUT when you want to shift a 500 M file, it gets over-saturated, and so a bit pointless... any of these tests on LightPeak??
Mean while external SATA connectors are the big solution to shifting big data... many MSs and cases have them now, so you just plug it in (hot plug too!) and get native speed and no problems!
"The beauty of Light Peak is that, if you pardon the expression, it's one cable to rule them all. It also only requires two components - the optical transport and the power feed rather than a cumbersome, fragile, multi-pinned, which-way-is-up?, interference prone connector and limited length cable."
From my point of view, it needs 2 cables rather than one. I think Intel etc. are forgetting about memory sticks, keyboards, mice, and almost everything people hook up to notebooks. I don't want to have to run a seperate power cord to my keyboard! I just replaced a bulky 250GB external HD with a slightly less bulky one that doesn't need power -- i for one will not go back to one that DOES need power. If Intel truly wants a universal cable they will have to have power over it.
you seem to have missed the point! A USB cable has FOUR wires.. 2 for power, 2 for 'balanced' data signal..
I am guessing Light Peak will just have the optical connect, plus power in ONE cable... and Its not impossible to have a couple more wires for 'legacy' USB connectivity..
as said above, none of this really matters, more than 'will people use it' .... we already have USB, ESATA, RJ45/Cat5, firewire(yes, sony and camcorders still use this!!), various optical stuff...
The real test, is will it work when your whole OS crashes, and your optical disc wont read???
- That is why most tech's want a floppy included! - do note that 'USB booting' is VERY new..
Biting the hand that feeds IT © 1998–2019