Slashdot Mirror
Intel Connects PCs To Devices Using Light
CWmike writes "Intel is working on a new optical interconnect that could possibly link mobile devices to displays and storage up to 100 meters away. The optical interconnect technology, Light Peak, could communicate data between systems and devices associated with PCs at speeds of up to 10Gbits/sec., said David Perlmutter, vice president and general manager of Intel's mobility group. The technology uses light to speed up data transmission between mobile devices and connected devices like storage, networking and audio devices, the company said. The technology could help transfer a full-length Blu-ray movie in less than 30 seconds, says a post on Intel's site. Light Peak can run multiple protocols simultaneously over a single cable, enabling mobile devices to perform tasks over multiple connected devices at the same time. 'Optical technology also allows for smaller connectors and longer, thinner, and more flexible cables than currently possible,' according to the Intel entry. It could also lead to thinner and fewer connectors on mobile devices, Perlmutter said."
What I've wanted for some time is a universal standard of structured cabling: I'd run a "bus" cable round the house, and in each room or termination point I'd have a box that allowed me to run different signals and different protocols over that bus - audio, HD video, ethernet, etc. No more running new cable runs each time I wanted to add a phone point, or an extra network socket. If this provides a way of doing this over a universal optical bus, then count me in...
So do I. It's called a "Blu-ray disc".
Anyway, when did "full-length Blu-ray movie" become a unit of data? What happed to the traditonal "Library of Congress" measure?
Having RTFA I am still at a loss to see how this differs from current 10Gb/s fibre optics. Is it just that they've given it a new name, as that's all that I can get out of the article.
I'd be interested in the cabling and connectors. 10Gb/s over fiber is certainly good, and would have a variety of fun uses; but is hardly groundbreaking, you've been able to get 10Gb over fiber for a while now.
To be putting it in consumer electronics, though, you pretty much have to make the cabling and connectors quite durable and generally idiot proof. This hasn't, historically, been the first set of attributes you associate with optical fiber(it's a hell of a lot more durable than you'd expect a tiny thread of glass to be; but you have to care about turn radius, and dust and stuff getting on the connectors, and whatnot). Either Intel is just handwaving, or they actually think that they've got a set of mechanical designs that'll let fiber be as robust as USB, and still work despite accumulations of pocket lint, and people rolling over cables with chairs, and stuff getting bent in laptop bags, and whatnot.
1) The article is about a cable.
2) You probably don't have a 10Gb/s cable
3) You certainly don't have a 100m long 10Gb/s cable.
# cat
Damn, my RAM is full of llamas.
Can anyone enlighten me as to which part of this story is meant to be news?
I am TheRaven on Soylent News
It's done over a cable - something the summary and most articles I've seen on it has failed to make apparent.
I did wonder about its usefulness myself, though. Why would I need to connect my iPhone to five different things at once? I rareky even need to connect my laptop to more than one or two things at a time.
Then I gave it some more thought and it occurred to me: at some point in the not-too-distant future smartphones will have the capabilities of today's laptops in terms of computing power and storage. You're unlikely to use that much power on the go, and you're hampered by the small screen and keyboard. But, for at least a segment of the population, you'll be able to dock your supersmartphone much like you can dock a laptop today. The dock will connect to a larger monitor, perhaps a keyboard and mouse (though those may be wireless direct to the phone), network, optical drive, offline storage, printer, and other peripherals. Your smartphone would be the computing guts of a much broader and capable system.
But the docking connectors on dock-able laptops are enormous compared to the size of a smartphone. Having a single, small, optically-based connector that can connect your phone to all those other devices will be key to this paradigm.
That is, of course, unless wireless technologies completely supplant wired connections for peripherals.
Source: http://techresearch.intel.com/articles/None/1813.htm - interesting facts
I'm confused as to how useful this is - can someone convert blu-ray movies to libraries of congress for me?
You could call it "S/PDIF"...
Yes I do, and Yes I do.
I have a bundle of at least 16 100Gb/s cables that run over 2Km. the only thing not letting my fiber optic cable run 160Gb/sec is the transceivers at each end are too low of quality to do so. so we live with 2 paltry 100Bt fibers a couple are used for video, and the rest are dark for future use.
This cable was laid 5 years ago way before Intel decided to discover fiber optics.
Do not look at laser with remaining good eye.
you haven't touched fiber for a decade have you. The new stuff can get bent to nearly an inch radius without even starting to suffer losses. Hell we have some jumpers her that were demoed to us that you can bend at a tight 110 degree angle with a ..5 inch radius and it still does not break, but does suffer from 2db loss at that point.
The bitch of fiber is that it's a PITA to install ends. I gave up and simply cut pre-made jumpers and fusion splice them onto the incoming. faster, cheaper, and far more reliable.
Do not look at laser with remaining good eye.
Is it just that they've given it a new name, as that's all that I can get out of the article.
So the non-article-reading crowd wins again. I gathered this much from the summary.
Well, the title was not very helpful - it came from the first of the linked articles. The second was a bit more informative but still quite vague.
The interesting thing here seems to be that they're planning to tunnel multiple protocols over the optical link. So you might be hanging monitors, USB devices, SATA drives, whatever off this link. It'd be a bridge that could tunnel your device connections to somewhere quite physically distant, using only a single cable. One assumes (maybe this is a big assumption) that an important part of the effort is in getting hardware that can efficiently do the encapsulation / decapsulation of the various device protocols. I'm not entirely sure why you couldn't do this over a 10Gb ethernet link, with some kind of protocol for tunneling over ethernet. I'd speculate that it'd make the controller chips more expensive if you did this but I really don't know. Everything is guesswork anyhow, until they give us more information.
The main thing I can see this being useful for is stuff like blade desktops - the real computer you're using as your desktop is just a blade server in a chilled room, with sysadmins leaving it regular sacrificial offerings for optimal uptime. The monitor, USB devices, everything would then be connected to the blade desktop by a single optical cable. Only one slim cable to route for each desktop, everything runs over it so the "desktop" can still have functional USB ports etc. Having an optical cable seems like it would be ideal for that kind of scenario. The ultimate thin client. If you have multiple Light Peak ports on a single blade then perhaps you could get multiple virtual machines to drive separate workstations, making your datacentre density even higher.
Other stuff it might be interesting for is some kind of cheap (?) high speed networking, home media servers, low cost SAN hardware, etc. Depending on how they do it of course. But if they made it generic enough it would be really interesting for a lot of applications that are now priced out of the reach of individuals and probably also small businesses.
Is this just cheap components for Fiber? 100 meters is pretty far, I am guessing that this could have networking uses beyond ripping media to external drives.
100m is a good distance... More than I'd probably need for connecting a mobile device to anything else in my house... But it isn't amazing. Doesn't good ol' ethernet cap out around 100m?
"Work is the curse of the drinking classes." -Oscar Wilde
Why hasn't some enterprising inventor come up with a cable/connector that combines optical (for data) and copper (for power) in a single cable?
Probably wouldn't be great for long distances, but I could imagine something like that having some advantages for replacing USB and ethernet w/PoE (at least in a home or office setting).
You can accomplish anything you set your mind to. The impossible just takes a little longer.
I think they got their bits and bytes mixed up. 10^13 bytes = 8*10^13 bits = 8000 seconds (2h13m) at 10^10 bits per second.
1 LOC is 2000 BRM.
The speed is 50 libraries of congress per microfortnight.
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
OMG! You can use light to transmit data over a cable? That's freaking crazy!! Wow.
What's next? Some way to switch circuits without using tubes or relays? Yeah -- like that would ever happen.
Actually, if you do your cabling right, yes you can. Ethernet's got distance limitations- fiber has less of one. Power can be ran the same distances if you pair it up around the fiber and make it part of a special connector... Moreover, the crowd they're tailoring this to doesn't care as much about power concerns over the interconnect. They want reliability, ease of cabling, distance, and overall speed- and they're not wanting to dangle all sorts of things like people do with USB stuff.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
...unless, of course, you take into account that S/PDIF is a protocol (like TCP, IP och Ethernet) that has nothing to do with the medium on which it's transmitted. You could have two monkeys yanking a rope (which does seem to be the case for the main internet-bearing lines accross the Atlantic from time to time) transmitting TCP/IP-packets between eachother.
Slightly thinner hairs?
... and then they built the supercollider.
I wondered how long a "full length Blu-Ray movie" is? Is it, like, just under 100 metres so it fits in the cable? Or is it 3 km, so that you have to drag that 100m cable for 30 seconds at 1 m/s to transfer it?
All these new units of measurement get me really confused.
Musicians don't die. They just decompose.
dust and stuff getting on the connectors, and whatnot .. and still work despite accumulations of pocket lint
Although this isn't mentioned specifically in the video, it appears as if the transceiver is meant to be permanently attached to the fiber. This would be the easiest solution to the lint issue, plus it would eliminate the complexity of making good optical connections. Essentially, I think they intend to have the transceivers molded into both ends of the fiber and it would probably look just like an USB cable to the average user, only with fiber running end-to-end, rather than copper. Of course, I'm not sure USB can reach 10 Gbs, so it probably would have a different type of electrical connection to the host PC.
I think the key innovation here is that they can have a short, high-speed electrical connection between the PC and the transceiver, and a large arbitrarily long fiber link between the transceivers themselves.
Do not stare into cable with remaining eye.
Liberal? Conservative? Compare perspectives at Left-Right
Hahaha... you're right. They must have used a float to calculate this... damn those Intel rounding errors!
Intel, try googling before you run: http://lmgtfy.com/?q=10+terabytes+%2F+10000000000+bps
Today, 1000baseT is included on $500 laptops, and you can get a 5 port 1000baseT switch for $25. If you think similar things won't happen with 10G, you're wrong.
No. with 10GBASE-T over COPPER it's not a question of COST, it's a question of POWER.
10GBASE-T uses too much power, all because it takes more power to get a higher bandwidth signal over the same 100m of copper as Gigabit. Current estimates are about 6W for a controller, which is way too high for integration into chipsets. And you can only reduce the power required for the DSP logic on the controller - the amount of power it has to deliver to the line to meet 10GBASE-T signaling standards is a fixed quantity. Since the power numbers I quoted are for second-generation devices, you can bet your ass that the DSP power quotient has been reduced to almost nothing, so that 6w is almost entirely signaling power.
Hell, even low-power Gigabit controllers use over a watt - this is why a lot of low-power devices still ship with just Fast Ethernet. The industry is moving toward EEE (Energy Efficient Ethernet) to combat this, but it's really nothing more than an idle mode - peak power is unchanged.
Optical offers very low power consumption at 10G (less than 1w), but will require the adoption of a new connector standard regardless of what standard the market goes with. Intel sees the opportunity here: if people will have to buy a new infrastructure to use low-power 10G, they will buy whatever is cheapest, not whatever is used in the server room.
Man is the animal that laughs.
And occasionally whores for Karma.