Light Peak to debut using copper instead of fiber optics

By on December 14, 2010, 8:00 AM
Intel has been touting Light Peak as the next big thing in connector technology and a way to reduce the proliferation of ports on modern computers, by using a single interface to handle everything from network connectivity to data transfers and transmitting high definition video to external displays. The interface can reportedly hit 10Gb/s transfer rates -- or twice the speed of USB 3.0 -- with the potential ability to scale to 100Gb/s in the next ten years. Indeed the company has showed some pretty impressive performance at several public events this year, but now it seems that “practical realities” have the company looking at more conventional technologies for its initial release.

Specifically, it seems Light Peak will debut using copper wires for the transmission of data instead of the technology's signature fiber optics. According to Cnet, however, the shift away from optical networking to traditional electrical interconnections won't affect the planned speed, which is still set at 10Gb/s bi-directionally. Intel hasn’t confirmed or denied the reports but if true it remains to be seen how this affects plans to eventually hit 100Gb/s.


Despite the move to copper cabling, Light Peak is said to be on track for a commercial rollout in the early first half of 2011, with Sony and Apple expected to be among the first to use the technology in its products.




User Comments: 29

Got something to say? Post a comment
cardriverx said:

OK, now I went from very intrested in lightpeak to not at all. Why would they even release LIGHTpeak with copper wire? Why not just wait longer? We don't need more connectors and cables unless it is actually a new tech - copper wire is not.

Guest said:

We don't need this. It will mean that the initial light peak using copper wires will become obsolete once they shift to fiber optics. It means more cables will go to waste. Just go to fiber optics directly so consumers can re-use those fiber optics all the way when 100GB/s becomes available.

Vrmithrax Vrmithrax, TechSpot Paladin, said:

I'm a little concerned by this as well. Does this mean we'll have stop-gap copper wiring for the beginning of LightPeak (which should really be called CopperPeak), and then all of those first run cables will become obsolete and need replaced when the real optical deal hits? Or is there some optical-to-copper converter on each end of the copper cables, so that they can just be easily replaced with the fiber cables later? Anyone have a definitive answer? Seems all I see when searching is vague answers and convenient skirting of forward/backward compatibility issues.

TorturedChaos, TechSpot Chancellor, said:

Very not thrilled by this news. Was kinda exciting when it was supposed to be fiber optics. But switching to another copper cable just sounds like a royal pain. I wasn't that thrilled about switching everything to a new cable any ways, but at least when it was fiber it was going to take us away from dependency on copper cables.

And I have to wonder too about compatibly when they finally release fiber cables. Now we are going to need to buy all new cables? New addon cards? WTH Intel......

Guest said:

Haha, so intel are going to "reduce the proliferation of ports on modern computers" by introducing a new port.... that is a stop gap...... that is based on current technology.....

Check it out, new copperpeak, transfer information over copper wire, its all new, the victorians were not doing that with the invention of the telephone or anything, honest.

Guest said:

So, Intel delays implementing USB 3.0 because they are coming out with the fantastic new technology and ends up with copperpeak. Just give me USB 3.0 and I'll keep waiting for true lightpeak.

Guest said:

I'm convinced that a marketing wog was responsible for this :-|

Cota Cota said:

Since i knew of Light Peak it has come to my mind, do we really need that desperate Light Peak? i mean im up from changing my Lan cable wich is incredible annoying to place when managing a server or using the non flexible cables, theres where Light Peak will show off doing a better job. However do we need that huge speed in our homes? what diveces could possible use all that speed?. For me it will be my SSD and thats about all, even my sata 3 HDD will be slacking (for me HDDs are for pure storage atm), its like having a end generation PC to only use Office, we have more gaps on our devices, not on our interfaces.

dividebyzero dividebyzero, trainee n00b, said:

I'm a little concerned by this as well. Does this mean we'll have stop-gap copper wiring for the beginning of LightPeak (which should really be called CopperPeak), and then all of those first run cables will become obsolete and need replaced when the real optical deal hits?

1. Copper wired means that the link can carry power as well as signal. IIRC most knee-jerk anti-Light Peak (when the tech was unveiled to the mainstream) comments were based upon needing a seperate power source for light peak products, so this in effect is a bridge between existing copper cables and peripherals and future LightPeak/TOSLINK products. LightPeak is already afaik aimed at IEEE P1901 and PoE compatibility

2. LightPeak can use existing TOSLINK cables ( see here [link] ), likewise the cable can also have HDMI, Display Port, Firewire and USB 3.0 connects (and can be utilised as such) - the only difference is likely to be cost since a fibre optic cable with copper plugs would need a transducer incorporated to convert the optical signal.

I'll now turn the forum thread back over to the Guest trolls and Flat-Earthers

Vrmithrax Vrmithrax, TechSpot Paladin, said:

The 100 Gb/s optical "jack-of-all-trades" interface that could be used for just about any kind of data connection, THAT was what made LightPeak so interesting. While I realize this slower copper version is a stepping stone in between now and getting to that 100 Gb/s holy grail, it's hard to justify switching everything over when it's marginally better than the widely adopted and available USB 3.0 standard.

I'm afraid that all of the delay tactics in supporting USB 3.0 in the chipsets, the chest beating and touting how superior LightPeak is, and the general ignoring of consumer demand for 3.0 has already put Intel behind the 8-ball for many. This rather lame launch of their "superior technology" may just turn out to be a bit of a dud.

dividebyzero dividebyzero, trainee n00b, said:

...it's hard to justify switching everything over when it's marginally better than the widely adopted and available USB 3.0 standard..

I doubt whether it's an either/or scenario. You're basing this on previous examples?

USB 2.0 introduction meant USB 1.1 and Firewire support no longer included on motherboards?

SATA 6Gb introduction meant SATA I/II and PATA support no longer included on motherboards?

HDMI and DP introduction meant D-Sub15 and DVI support no longer included on motherboards and VGA's ?

PCI Express introduction meant PCI support no longer available on motherboards?

I'm afraid that all of the delay tactics in supporting USB 3.0 in the chipsets

By all means please provide a convincing argument¹ on how having extra PCI lanes available to the DMI or QPI for USB 3.0 substantially improves on the feature set, pricing and bandwidth offered by a third party controller.

(I'll narrow it down a little: the controller is a sub $2 item and is 10mm x 10mm x 0.65mm in size)

the chest beating and touting how superior LightPeak is.

SATA I (at 150Mb/sec) offers little real-world improvement over IDE (at 133Mb/sec)- how did that work out?

and the general ignoring of consumer demand for 3.0.

Excluding the obvious Anti-Intel trolls *ahem* where is this demand for USB 3.0?

Most external drives have powered eSATA options (as do most motherboards)...are you telling me that eSATA is inferior to USB 3.0 ?

If USB 3.0 was in such demand, don't you think that Razer et al wouldn't have jumped on the bandwagon offering mice and keyboards, webcams, coffeemug warmers etc... There are certainly enough USB 3.0 equipped boards available, and even a number of chassis offering USB 3.0 connectivity.

And of course, the majority of users are so sensitive to bandwidth yet wouldn't have a clue whether their Gb LAN is PCI or PCI-E based, nor bother to check whether the SATA device they are plugging into a board works through the I/O hub or a third party controller* and the difference in read/write performance

has already put Intel behind the 8-ball for many.

Then I guess Intel is doomed...oh well, it was fun while it lasted

¹ If I'm convinced of the soundness of the logic I will personally write to Dirk Meyer and recommend you for the Gold Star (1st Class) or Employee of the Month (whichever is applicable)

*Guess what percentage of systems I check/upgrade that have hdd's/odd's plugged into a port that "looks the least messiest" with regards cable routing as opposed to offering the highest data throughput.

Guest said:

I'm assuming what happened is that USB 3.0 was nearing commercialisation while LightPeak was still a half-baked project, Intel's being doing everything it can to slow down USB 3.0 adoption by not supporting it in their chipsets, while hyping LightPeak as the next big thing, and now with USB 3.0 stating to get a foothold on the market before LightPeak is ready, they're resorting to this stopgap measure to prevent it becoming another FireWire.

I still hope LightPeak does work out, and proves to be all it says it is. Ditching PS/2 connectors and parallel ports for USB 1.0 was great at the time, ditching HDMI, USB 2.0/3.0, RJ45, SATA, eSATA for a single interface would be great now, can't see it happening though,

Vrmithrax Vrmithrax, TechSpot Paladin, said:

dividebyzero said:

I'm a little concerned by this as well. Does this mean we'll have stop-gap copper wiring for the beginning of LightPeak (which should really be called CopperPeak), and then all of those first run cables will become obsolete and need replaced when the real optical deal hits?

1. Copper wired means that the link can carry power as well as signal. IIRC most knee-jerk anti-Light Peak (when the tech was unveiled to the mainstream) comments were based upon needing a seperate power source for light peak products, so this in effect is a bridge between existing copper cables and peripherals and future LightPeak/TOSLINK products. LightPeak is already afaik aimed at IEEE P1901 and PoE compatibility

2. LightPeak can use existing TOSLINK cables ( see here [link] ), likewise the cable can also have HDMI, Display Port, Firewire and USB 3.0 connects (and can be utilised as such) - the only difference is likely to be cost since a fibre optic cable with copper plugs would need a transducer incorporated to convert the optical signal.

I'll now turn the forum thread back over to the Guest trolls and Flat-Earthers

Ahhh, finally a reason that makes sense for the copper! Everywhere else you look, this story is laid out like Intel just said "oh by the way, you don't get optical, deal with copper" and never really states any reasons behind why.

Makes much more sense to me now... Thanks for the bit of clarity in an otherwise murky subject, @dividebyzero... I hereby rescind my declarations of LightPeak's copper launch "lameness" and move to a more reserved "now let's see if it gets adopted" stance. I really do hope to see it flourish - seems to have much more potential as a truly "universal" interface, particularly when those transmission speeds approach the goals Intel has laid out.

lawfer, TechSpot Paladin, said:

dividebyzero said:

...it's hard to justify switching everything over when it's marginally better than the widely adopted and available USB 3.0 standard..

I doubt whether it's an either/or scenario. You're basing this on previous examples?

USB 2.0 introduction meant USB 1.1 and Firewire support no longer included on motherboards?

SATA 6Gb introduction meant SATA I/II and PATA support no longer included on motherboards?

HDMI and DP introduction meant D-Sub15 and DVI support no longer included on motherboards and VGA's ?

PCI Express introduction meant PCI support no longer available on motherboards?

I'm afraid that all of the delay tactics in supporting USB 3.0 in the chipsets

By all means please provide a convincing argument¹ on how having extra PCI lanes available to the DMI or QPI for USB 3.0 substantially improves on the feature set, pricing and bandwidth offered by a third party controller.

(I'll narrow it down a little: the controller is a sub $2 item and is 10mm x 10mm x 0.65mm in size)

the chest beating and touting how superior LightPeak is.

SATA I (at 150Mb/sec) offers little real-world improvement over IDE (at 133Mb/sec)- how did that work out?

and the general ignoring of consumer demand for 3.0.

Excluding the obvious Anti-Intel trolls *ahem* where is this demand for USB 3.0?

Most external drives have powered eSATA options (as do most motherboards)...are you telling me that eSATA is inferior to USB 3.0 ?

If USB 3.0 was in such demand, don't you think that Razer et al wouldn't have jumped on the bandwagon offering mice and keyboards, webcams, coffeemug warmers etc... There are certainly enough USB 3.0 equipped boards available, and even a number of chassis offering USB 3.0 connectivity.

And of course, the majority of users are so sensitive to bandwidth yet wouldn't have a clue whether their Gb LAN is PCI or PCI-E based, nor bother to check whether the SATA device they are plugging into a board works through the I/O hub or a third party controller* and the difference in read/write performance

has already put Intel behind the 8-ball for many.

Then I guess Intel is doomed...oh well, it was fun while it lasted

¹ If I'm convinced of the soundness of the logic I will personally write to Dirk Meyer and recommend you for the Gold Star (1st Class) or Employee of the Month (whichever is applicable)

*Guess what percentage of systems I check/upgrade that have hdd's/odd's plugged into a port that "looks the least messiest" with regards cable routing as opposed to offering the highest data throughput.

This is what I call "owning someone with class." Absolutely loved the little footnote too.

Guest said:

"Indeed the company has showed some pretty impressive performance"

corrected typo should read as follows:

has shown

Thank you,

TrueBooleanFalse

princeton princeton said:

What they should try to do in the future is use both copper and fiber. Would putting that in the same wire work if it had some sort of insulator between them? That way it could transfer power over the copper and data over the optical fibers.

dividebyzero dividebyzero, trainee n00b, said:

That's basically the idea they are going with. Copper for power delivery, fibre optic for data transfer.

Guest said:

y not InfiniBand? it's far more superior then Light peak or any others available buses... well except ARCH(developed in china provide 2x the speed of InfiniBand and currently used in TianHe-1A)

honestly i dont see any "advance" thingy in this thing....and how it replace USB since it dosent provide power to device? @dividebyzero using copper wire dosent mean it carry power to support device tat connected it

no offence jus curious.

InfiniBand http://en.wikipedia.org/wiki/InfiniBand

dividebyzero dividebyzero, trainee n00b, said:

@dividebyzero using copper wire dosent mean it carry power to support device tat connected it

[link]

need any more hints?

Guest said:

"dividebyzero said

a fibre optic cable with copper plugs would need a transducer incorporated to convert the optical signal.

I'll now turn the forum thread back over to the Guest trolls and Flat-Earthers"

Get a clue mate, if you are turning the signal from copper>fiber>copper there is no point in the fiber at all as it limits both ends at the copper. Plus do you know how big a transducer would be on every cable? And the cost of said cable? Plus would you like to tell me how having a transducer at the end of a cable would power a portable hard drive, like you said copper can carry current but fiber cant, is it a magical transducer?

Now stop being so aragant as to think having a user name on a forum makes you better than a "guest", there are always going to be people smarter than you.

PanicX PanicX, TechSpot Ambassador, said:

Guest said:

Get a clue mate, if you are turning the signal from copper>fiber>copper there is no point in the fiber at all as it limits both ends at the copper.

There are qualities a fiber media provides that are an improvement over copper, e.g. attenuation, crosstalk, RF interference. How much of this applies to lightpeak, I'm not sure, but going full copper would be a mistake as the benefits of fiber data transition is the only worth while improvement I see in this technology.

Guest said:

We have 100 Mbps and even 1Gbps communications on Ethernet cables now that are copper. There is even 10Gbps Ethernet on Enterprise sized hardware. Why are we still struggling to get a crummy10 Mbps on LightPeak whether copper or optical? I must be missing something here. --Steve

Guest said:

The main problem is not the media, it is all the company that use different end connectors. Just at home I have 5 different USB cables because each device I have come with a different end connector. Start by using all the same connector and at the same time all the same power connector and the world will be much better ;)

Vrmithrax Vrmithrax, TechSpot Paladin, said:

Guest said:

We have 100 Mbps and even 1Gbps communications on Ethernet cables now that are copper. There is even 10Gbps Ethernet on Enterprise sized hardware. Why are we still struggling to get a crummy10 Mbps on LightPeak whether copper or optical? I must be missing something here. --Steve

Consider that LightPeak will be able to do Ethernet communications AND display/media connections AND drive interfacing AND general peripheral connections... Simultaneously, by the way...

Ethernet cables see a fraction of the signal action that LightPeak will be dealing with, and those Ethernet signals are in nice tidy packets, not the flailing melee of signals and data LightPeak will be dealing with It's an apples-to-oranges comparison. Now compare LightPeak to USB types of interfaces, and you're closer to relevance.

dividebyzero dividebyzero, trainee n00b, said:

Get a clue mate, if you are turning the signal from copper>fiber>copper there is no point in the fiber at all as it limits both ends at the copper..

copper>copper>copper for power delivery

copper>fibre>copper for data transmission

What part of that did you not understand?

Now stop being so aragant (sic) as to think having a user name on a forum makes you better than a "guest", there are always going to be people smarter than you.

Thats a given...although I doubt, judging by your comprehension, grammar and spelling skills that you are one of them.

captaincranky captaincranky, TechSpot Addict, said:

Might I be so bold as to remind everyone that electricity travels at or near the speed of light anyway.

So, let's play "rock, paper, fiber" or maybe that was, "copper scissors, glass".

So when light peak is fully operational, will an email cause the light bulb on top of my head to come on?

dividebyzero dividebyzero, trainee n00b, said:

Might I be so bold as to remind everyone that electricity travels at or near the speed of light anyway.

Ah, but you're talking about regular electricity I take it? How about the reaaaaaalllly slow kind of electricity that is generated in the neurons of fanboys, trolls and Baltimore Ravens fans?

So when light peak is fully operational, will an email cause the light bulb on top of my head to come on?

Please contact your ISP for the latest equipment upgrade.

Guest said:

InfiniBand has my vote.

"But, if we give everyone InfiniBand then how will we maintain control?"

Guest said:

Yawn....

When Intel decides what they have and how they are going to release give me a call.

They've changed so many times what it's going to be and what it's going to do.

In the meantime, Intel, please put USB3 on your chipsets.

Load all comments...

Add New Comment

TechSpot Members
Login or sign up for free,
it takes about 30 seconds.
You may also...
Get complete access to the TechSpot community. Join thousands of technology enthusiasts that contribute and share knowledge in our forum. Get a private inbox, upload your own photo gallery and more.