Apple and Intel Collaborate on Next Generation 'Light Peak' Connectors?

[twilson]

Well, if this comes out in 2010, it'll make SATA3.0/eSATA/USB3/FW3200 all redundant before they get mainstream.

It might get a tad confusing if every connector on the computer has the same shape.

I think it's more of a way to aggregate lots of devices and ports into a hub and a single connector.

As a result this will reduce the amount of space taken up by the current connectors allowing the laptops to be made smaller, as it is the physical connector sizes that are preventing netbooks/laptops/Macbooks from getting much smaller at the moment.

 

[chrisdee]

Finally som happy rumors

 

[sammich]

That's exactly what I said, to justify why people where discussing the issue ...

Doh.

I think it's more of a way to aggregate lots of devices and ports into a hub and a single connector.

As a result this will reduce the amount of space taken up by the current connectors allowing the laptops to be made smaller, as it is the physical connector sizes that are preventing netbooks/laptops/Macbooks from getting much smaller at the moment.

I was thinking...you can't really get any smaller (read: thinner) than the current macbooks, then I thought it would really enable current form factors to take on many more connectors within the same limit of space. That would mean the MBA could be equipped with like 2/3 (or 4) LP connections.

 

[toke lahti]

Macrumors, please do something about the 'Sunday off' situation! Perhaps, you could get somebody to write a few little pieces and control the comments.

Yes, this is horrible!
People talking about rumored technologies or even existing ones in MacRumors Forums!
If this is why MacRumors exist, maybe they should shut down the whole site!

 

[djellison]

LP for Mice.

Yeah - that'll be great.

 

[toke lahti]

Well except in Europe where Apple signed on to standardization of micro-usb connections.

I guess all phones will have micro-usb in addition to LP. People are not happy to carry dongles, that cost almost as much as the phone, with their phones.

 

[toke lahti]

As a result this will reduce the amount of space taken up by the current connectors allowing the laptops to be made smaller, as it is the physical connector sizes that are preventing netbooks/laptops/Macbooks from getting much smaller at the moment.

Size of a laptop is more defined by screen and keyboard size. If there comes foldable or scrollable screens to market, then size can decrease.

So what will really happen to MB/MBP with LP?
Usb & ethernet have to stay, otherwise they loose too many customers.
MiniDP changes to LP and firewire changes to LP and you'll have to buy dongles for each.
If you need to connect any 3rd peripheral (2nd monitor, hdd, etc...) you'll need LP hub with dongles, which will cost about $500...

 

[NintendoFan]

Size of a laptop is more defined by screen and keyboard size. If there comes foldable or scrollable screens to market, then size can decrease.

So what will really happen to MB/MBP with LP?
Usb & ethernet have to stay, otherwise they loose too many customers.
MiniDP changes to LP and firewire changes to LP and you'll have to buy dongles for each.
If you need to connect any 3rd peripheral (2nd monitor, hdd, etc...) you'll need LP hub with dongles, which will cost about $500...

I think when he's talking about "smaller" he means "thinner".

 

[Macinposh]

So what will really happen to MB/MBP with LP?
Usb & ethernet have to stay, otherwise they loose too many customers.
MiniDP changes to LP and firewire changes to LP and you'll have to buy dongles for each.
If you need to connect any 3rd peripheral (2nd monitor, hdd, etc...) you'll need LP hub with dongles, which will cost about $500...

Usb and ethernet might stay in the first rev,after 2nd/3rd when the LP has gained ground,they will drop them in favour of a second/third LP.

For the hub: The most probable thing is that dispays are going to act as a "MainHub" with 4-6 LP ports. There you can connect whatever.
Intel is probably pushing for wireless electricity for the mouse/keyboard at that time,so you would get rid of the extra clutter as well.

My guesses.

 

[Erasmus]

No one is arguing that fiber isn't fast, all I was trying to say is at 10Gb/sec can easily be done with copper today and with copper you can still power/charge a device. In this case length isn't that big of a deal because this is a desktop standard. Well I guess if you own a house like Bill Gates you might have a problem with copper.

And I'm trying to say that it can't. If you could do it "easily", USB would have been there by now. I know Wikipedia is a dodgy source of information, but it says even the superspeed transmission in USB 3 of 4.8 Gbit/s is limited in transmission distance to a few metres. That's fine for most desktop applications, but Light Peak is about replacing display connectors, and even ethernet - So no home cinema or high-speed LAN for you. Scale it up to and past 100 Gbit/s, and suddenly the range of electrical transmission is just centimetres. And that doesn't fit well with your "desktop standard". Optically however, transmission distance remains irrelevant.

Throwing two DC copper wires into the cord to charge a device is an obvious solution to charging that has no drawbacks. Still no shielding required, as they are only there to transmit power, not data. Even if the extra lines are there just to flesh out an otherwise extremely thin and fragile cable.

 

[Erasmus]

Imagine this, you have 5 devices hooked up to your computer.
The devices are resting on your desk, the computer is hiding under the desk, together with all the cables.
How to find the cable of your keyboard?
Hence it would be nice to know which cable belongs to which device.

Just a simple clarification

Same way that it has always been done.

Ingredients:
• One roll of masking tape
• One black pen.

Instructions:
• Apply liberally.

 

[dernhelm]

I was talking with a this weekend friend that works at Internet2 about the 10Gbs Ethernet standard. He indicated that the interesting implementations of it were accomplished using fiber cables, and that copper was simply too restrictive (quality/length/cost) for standard use. They were a little discouraged at its slow adoption rate (they're a research organization working with dozens of universities, but it always helps when the tech you are promoting/researching takes off).

This was all before I even read this article, but he said he thought 2010 was going to be the year that 10Gbs Ethernet started hitting the upward slope of the adoption curve. He said it would be due to some announcements by hardware vendors of support and less expensive components. At the time, I thought nothing of it, everyone wants more speed, and component prices always drop. But now I wonder if he wasn't really referring to technology like this.

 

[whooleytoo]

I wonder if there's any risk in this. If you plug a monitor into your monitor port, you know it's going to send a video signal. If you plug a router into your Ethernet port, you know it's going to send & receive network data.

But if you plug a device into a Light Peak port, you don't know how that device is going to identify itself.

 

[doctor-don]

Kinda underwhelmed...

Yes, it is faster speed, but the cabling/connectors look quite pricey. For the comparisons used by Intel in the Youtube video linked here, optical cable is not cheap.

We're likely looking at fiber channel networking, maybe the odd display (though Display Port hasn't taken off, and VGA is still quite prevalent.) But, the influx of cheap printer connectors, cheap camera connectors, and cheap mouse connectors. Firewire lost out to USB in a lot of respects simply due to need, and costs involved in implementing... meaning that if we aren't committed to changing all our peripherals over to ones that utilize expensive cables, we're stuck with machines attempting to maintain backwards compatibility.

My recent TOSLink cable cost me one penny plus $2.98 shipping for a 6 foot cable. I just ordered more cables at less than a dollar for a 12 foot cable and less than $2 for a 25 foot cable. Of course, the shipping is $2.98 FOR EACH CABLE, even when ordering 2 cables. It makes no sense, but it is still cheaper than the local store. These prices are lower than those for USB cables at the same site - about $1 per foot.

 

[pdjudd]

But if you plug a device into a Light Peak port, you don't know how that device is going to identify itself.

Presumably the device will identify itself to the CPU like standard USB devices do.

 

[doctor-don]

I wonder if there's any risk in this. If you plug a monitor into your monitor port, you know it's going to send a video signal. If you plug a router into your Ethernet port, you know it's going to send & receive network data.

But if you plug a device into a Light Peak port, you don't know how that device is going to identify itself.

Same way any USB device identifies itself - external hard drive, flash drive, SDHC/CF card, mouse, keyboard, whatever.

 

[PracticalMac]

Well, if this comes out in 2010, it'll make SATA3.0/eSATA/USB3/FW3200 all redundant before they get mainstream.

It might get a tad confusing if every connector on the computer has the same shape.

Yeup.

And it is an Intel+Apple cooperative by the looks of it, so it will have far more support then if either one did it alone.

You forgot, it will likely supplant GigNet as well in small (home office size) LANs.

This is rather a stunner.

 

[-hh]

Well, if this comes out in 2010, it'll make SATA3.0/eSATA/USB3/FW3200 all redundant before they get mainstream.

It might get a tad confusing if every connector on the computer has the same shape.

Um...no...

The whole point is that all the connectors are the same. You can plug anything into any port.

This Uni-Port dimension is pretty interesting. Its pretty trivial for a device to self-report what it is and how much bandwidth it expects, which means that it doesn't really matter for what gets plugged in where - - its not unlike having N USB devices hanging off a single hub: there's nothing special about USB Hub Port #0 versus the others that dictates a particular preference.

Similarly, Firewire has had the ability for "pass through" FW ports to daisy-chain (and/or hub). It just isn't seen/used as much.

Overall, I expect that this would work quite similarly, and the interesting part is that at each point where there's copper being tapped out, it is also possible to have a hub/router-esque interface, so a near-term implimentation could very easily be LP cable to an LCD monitor to drive its video, plus a couple of USB ports that are used for the keyboard, mouse, plus even a SATA port for an internal HD built into the monitor, and an external FW port to drive an external HD, etc, etc.

At face value, it sounds more like a reincarnated ADC (Apple Display Connector) than a replacement for USB.

Only in that it is advertising bandwidth sufficient for video?

The big question is why is Intel promoting USB 3.0 if they have Light Peak coming around the corner (2010). And why did Apple (who has never been afraid to develop a new standard) have Intel develop it?

I think the better question is if this is a move by Intel (with Apple's assistance) that is going to "game change" on USB3 before it gets a viable foothold. USB3 has been struggling for a variety of reasons, not the least of which includes the technical merits of its specific implimentation...we've seen a similar situation emerge with External SATA, and while eSATA is less bad in terms of connector quality, it doesn't resolve the "birds nest" cables situation, being that many home PC users are now in the situation of having to provide more plugs for a dozen vampire power bricks, which ends up being a bigger hassle than the original desktop.

Imagine running one fiber line to every room of your house. You could keep your computer in the closet and just have a keyboard, mouse and monitor hooked up where you want to use it!

or

You could have 1 insanely fast computer with tons of storage in the basement and use this to bring your own Video on Demand service to every tv and computer in your house!

I've had that same vision for several years now. I was able to test it partially around 8 months ago with VNC, but its utility was limited due to the 100bT Ethernet on the Powerbook G4 that I was using as a smart terminal.


-hh

 

[griz]

ONE port to rule THEM ALL !!!

sorry could not resist this.

Couldn't believe it took that long for someone to say it.

 

[doctor-don]

Finally

a Wikipedia article on Light_Peak - partially due to its being introduced September 23.

Advantages include a run up to 100 meters. Since most external devices require their own power sources, that is where the outlet / port should be located for what were formerly USB chargers.

 

[PaperMacWriter]

I dont think that Apple is looking to kill USB just yet.... I'm guessing that the short term goal(meaning that it's close enough to now to have a remote chance of guessing right) is to kill ethernet, FW, and Mini-DisplayPort. This would accomplish several feats for them:
1) remove the bottleneck in thickness that is ethernet
2) Remove the next bottleneck that failed to become popular enough that is FW
3) Possibly(I haven't seen its size yet, so I cant be sure) replace the iPod 30-pin dock connecter, or at least the computer end of it, causing force mass adoption(FW doesn't work with new iPods, forcing all other computers to support it. So if there was no LP (meaning no iPod and iPhone syncing), it could be a tipping point to get something else).
4) Shut us up about the MBA. This would at least replace the mDP, giving them one of the things that we have whined about: Wired networking. This way we'd all shut up and (possibly) buy one (at least over the competition).
5) Internal wiring. I could see this becoming standard as SSDs begin to seep over the 6.0Gb/s line(and I would imagine before that for Apple.... even if just for bragging rights...).
6) Replace HDMI. If I understand it right, this has greater bandwidth then HDMI, and seems to be superior. AND IF(and only if) this has no licensing fees, then I think that it will take over HDMI in a short period of time.

That's my guess for what this could accomplish that apple might like that isn't too far over the horizon... plans bigger than this, there is no point in guessing.

Oh, and if I didn't understand something about it right, my apologies. These are predictions to the best of my current understanding....
and if anyone has a size comparison to another port(or it's dimensions), could you post them please? I would love to see the port itself....

So that's what I see coming... eventually it will send USB the way of floppy, but that's not too soon.....

SG

 

[avionicsman]

Fiber optics, about time slow pokes!!

We have had fiber optics in aviation for a decade and half. Although with this technology maybe it will catch on even more in my world. Honeywell had used it in the HF communication boxes and although it was cool to see the setup and install was a bitch. Very fragile wiring and connectors. I hope these connection will be more robust.

 

[ss957916]

I'm concerned the demo shows a USB-type connector. They're too large and thick.

Also, am I the only person who wants a lockable connector?

 

[lkrupp]

I am glad Apple is thinking about cutting down on the cords. But, how exactly will my future iPhone charge with light?

There's nothing to prevent the final connector from including both optical fibers and a couple of copper wires for powering/charging the connected device. There's also the possibility of wireless charging which Apple and others are developing as we type (remember Nicola Tesla?).

 

[lkrupp]

I'm concerned the demo shows a USB-type connector. They're too large and thick.

Also, am I the only person who wants a lockable connector?

Yes, you would probably be the only one. A locking connector would be a disaster wating to happen. Why do you think Apple came out with the Magsafe power connector. A locking connector would mean laptops and devices crashing to the floor as well as broken (expensive) cables.