Intel: Light Peak interconnect is ready, but built on copper

Jos

Jos

Posts: 3,073   +97
Staff

Intel is greeting us with some good and bad news this week. In an interview with IDG News, the company’s executive vice president David Perlmutter said that Light Peak is ready to be implemented by device makers. But as rumored, its first implementation will use old copper wires for data transmission, instead of fiber optics, in order to save costs. "The copper came out very good, surprisingly better than what we thought," Perlmutter said.

The company would presumably still be able to hit its initial target of 10Gbps speeds using copper cabling which, according to Intel, will be more than adequate for the majority of user needs today. By comparison, the relatively new USB 3.0 protocol sports a maximum transfer rate of 4.8Gbps. That doesn’t mean that Intel isn’t working on a fiber optic connection. Eventually they will make the move to fiber optic cabling – with the potential of reaching up to 100Gbps transmission speeds after a few years – it’s just going to take longer for it to be implemented.


Permutter didn’t say which companies are the first supporters and when complete products will be shipped. Rumors have pointed at Apple and Sony as two of the main forces pushing Intel to get Light Peak out the door, so we’ll be looking forward to any announcements from them in the upcoming months.

Permalink to story.

https://www.techspot.com/news/41901-intel-light-peak-interconnect-is-ready-but-built-on-copper.html

 
Intel puts the brakes on USB 3.0 and tried to slow it's progress. Now we know why, they don't really have light peak ready, at least not how they envisioned. "We have this great fiber optic high speed technology, however it is not quite there yet and only runs on copper".
 
Guest said:
Intel puts the brakes on USB 3.0 and tried to slow it's progress. Now we know why, they don't really have light peak ready, at least not how they envisioned. "We have this great fiber optic high speed technology, however it is not quite there yet and only runs on copper".
Click to expand...

So you ignored the fact that light peak is over two times as fast as USB3?
 
Seems like they're making some interesting moves. LightPeak is awesome...but with copper. SandyBridge is awesome...but with no DX11. Both are going to require purchases of second generation hardware to reach its true advertised potential.

Not that both aren't technological advances, but for them to harp on these new technologies for over a year, and at the last second to announce "but..." seems like a black eye.
 
Princeton said:
Guest said:
Intel puts the brakes on USB 3.0 and tried to slow it's progress. Now we know why, they don't really have light peak ready, at least not how they envisioned. "We have this great fiber optic high speed technology, however it is not quite there yet and only runs on copper".
Click to expand...

So you ignored the fact that light peak is over two times as fast as USB3?
Click to expand...
Sometimes, just sometimes, backwards compatibility is important too. That, and the general population is usually much slower at embracing completely new (to them) technology/interconnects, so we'll see how it all works out.
 
madboyv1 said:
Princeton said:
Guest said:
Intel puts the brakes on USB 3.0 and tried to slow it's progress. Now we know why, they don't really have light peak ready, at least not how they envisioned. "We have this great fiber optic high speed technology, however it is not quite there yet and only runs on copper".
Click to expand...

So you ignored the fact that light peak is over two times as fast as USB3?
Click to expand...
Sometimes, just sometimes, backwards compatibility is important too. That, and the general population is usually much slower at embracing completely new (to them) technology/interconnects, so we'll see how it all works out.
Click to expand...

Still, we dont really need that huge speed, until we get faster devices like the SSD (wich is slow for the Light Peak), Light Peak is just a X-34 in the garage, we have a gap in the devices not in the interface.
 
Yah, I'm more than a little curious as to how this is going to play out... USB has backward compatibility, so you can still plug 2.0 hardware and cables into 3.0 sockets. So are you going to be able to plug all of your CopperPeak equipment and cables into LightPeak when it's actually ready?

I hope they don't intend for all of the early adopters will just buy everything new when they upgrade to optics... Could weigh heavily into considerations on whether to jump to the faster CopperPeak iteration now, or just stick with something like USB3.0 and wait for the uber-fast true LightPeak version later.
 
Cota said:
Still, we dont really need that huge speed, until we get faster devices like the SSD (wich is slow for the Light Peak), Light Peak is just a X-34 in the garage, we have a gap in the devices not in the interface.
Click to expand...

I think the extra speed will be more useful when, as you say, the faster drives are commonplace. But, it's also going to be useful for the other main target for LightPeak: media transmission. They have already shown live demos of a single LightPeak connection driving multiple HD video streams to displays, among other simultaneously connected peripherals, without a hiccup. That's where LightPeak could be truly amazing - a single connection could theoretically handle networking, display, and all peripherals for your PC. Excess speed would be useful for that kind of traffic.
 
gwailo247 said:
Seems like they're making some interesting moves. LightPeak is awesome...but with copper. SandyBridge is awesome...but with no DX11. Both are going to require purchases of second generation hardware to reach its true advertised potential.

Not that both aren't technological advances, but for them to harp on these new technologies for over a year, and at the last second to announce "but..." seems like a black eye.
Click to expand...

Intel never advertised DX11 for sandy bridge. You can't say full advertised potential if it wasn't advertised.
 
For Princeton{ A ton of prince?}:Don't you have your own ideas other then Intel and nVidia rule? Some of us are gettin' bored,Oh, sorry, you're just like Intel now!...no ideas, only steals...I mean improvements.
 
At 4.8MBps, USB 3.0 is way behind what Light Peak is reported as supporting. So my point being, who cares if USB 3.0 is done and gone already If Light Peak will be, in it's first release, over twice as fast, why do we care if USB ever gets any faster? The potential of Light Peak makes USB sound like old news. Sure it's not actually using fiber yet, but if copper is still twice the speed, something is wrong with USB and it needs to be replaced. Think of Apple-Talk. and how up in arms all those yahoos were when it was replaced. They got over it, their world didn't crumple and they are much better off now without it.
 
Princeton said: Intel never advertised DX11 for sandy bridge. You can't say full advertised potential if it wasn't advertised.
Click to expand...

I'm not Intel bashing, you're right, they may not have explicitly advertised it. I guess I just made an *** out of me by assuming that a game changing chipset is going to have support for a graphics standard that has been around for a few years. I know it doesn't make a difference for the average notebook or tablet consumer which is going to be using this chipset and will not need DX11 graphics, but I still feel that this is something that would hold me back from purchasing this chipset for my own needs.

I'm never an absolute fanboy of everything, I laud what works, I criticize things that I feel have problems. I feel that lately, starting with USB 3.0, now the previously mentioned issues, Intel has been focusing more on the marketplace than the product. And I don't like it when companies do that. So I'm going to ***** about it. That is not going to stop me from being a loyal Intel customer. =D
 
Guest said:
At 4.8MBps, USB 3.0 is way behind what Light Peak is reported as supporting. So my point being, who cares if USB 3.0 is done and gone already If Light Peak will be, in it's first release, over twice as fast, why do we care if USB ever gets any faster? The potential of Light Peak makes USB sound like old news. Sure it's not actually using fiber yet, but if copper is still twice the speed, something is wrong with USB and it needs to be replaced. Think of Apple-Talk. and how up in arms all those yahoos were when it was replaced. They got over it, their world didn't crumple and they are much better off now without it.
Click to expand...

Valid points, except comparing the massive widespread standard of USB to a tiny computer company's failed attempt at instituting their own version of networking really doesn't make sense. The number of Apple products (and users) affected when Apple Talk was shelved is tiny compared to the current number of USB appliance and peripheral users. It's nowhere close to comparing apples to apples. More like comparing a watermelon to a raisin. There is a huge pool of USB equipment and USB-enabled technology in the mainstream now, with more coming every day. LightPeak has a huge mountain to climb to just replace USB.

That said, would love to see a USB-connectivity option for LightPeak (like a connection hub or something similar). Would allow people to migrate from USB to LightPeak easily over time. More than likely, though, we will just have hardware with both USB and LightPeak connections for a while, until one of the standards wins the interface deathmatch. My money is on LightPeak, in the end. Provided they can achieve their vastly superior bandwidth goal, that is.
 
Vrmithrax said:
That said, would love to see a USB-connectivity option for LightPeak (like a connection hub or something similar). Would allow people to migrate from USB to LightPeak easily over time. More than likely, though, we will just have hardware with both USB and LightPeak connections for a while, until one of the standards wins the interface deathmatch. My money is on LightPeak, in the end. Provided they can achieve their vastly superior bandwidth goal, that is.
Click to expand...

This is pretty much floppy/cd drives all over again. Floppy was the way to go, then cd came in. For the longest time, machines had drives for both, and we are now seeing the final end of the floppy. This is what's going to happen with USB/LightPeak. Both will be integrated onto the same machine until USB fades out (which like the floppy, will take several years). Face it, Light Peak is the future. It's so much better than USB and has much more potential. It's just going to take some time to fully integrate. This will not be an "overnight" thing like people seem to think. Stop trying to hinder the future and look at the past...it will show you how the future's going to go.
IEEE 1394 is still trying to outdo USB, but lightpeak will come right out way ahead of USB which gives it the advantage. As USB improves, so will light peak until USB realizes it can't keep up. Poor IEEE 1394 just needs to give up, lol.
 
Nonsense, USB will never fade out, not into the realm of LightPeak or similar, because USB passes voltage to feed the devices as one key requirement, while LightPeak type of connection doesn't. In other words, USB is a full Plug'nPlay protocol, while LightPeak is for data exchange only. Thenceforth, they are in no competition.
 
Vrmithrax said:
Valid points, except comparing the massive widespread standard of USB.... compared to the current number of USB appliance and peripheral users....There is a huge pool of USB equipment and USB-enabled technology in the mainstream now, with more coming every day. LightPeak has a huge mountain to climb to just replace USB.
That said, would love to see a USB-connectivity option for LightPeak (like a connection hub or something similar)....
Click to expand...

Haven't we been here before?

dividebyzero said:
Vrmithrax 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?
Click to expand...
LightPeak can use existing TOSLINK cables ( see here http://www.youtube.com/watch?v=nfGevFIVKw4&feature=player_embedded), likewise the cable can also have HDMI, Display Port, Firewire and USB 3.0 connects (and can be utilised as such)....
Click to expand...

The Light Peak cable plugged into the laptop through a USB 3.0 port with components added to receive the optical signal. When asked if that would be the standard port used for Light Peak, Kahn said USB 3.0 is a likely place to start because it is common, but that "you could take the size way, way down". That could reduce the amount of space needed for a Light Peak port, a crucial consideration for small devices like handheld computers.

The laptop in the demo could still accept normal USB 3.0 devices in its USB port.
Click to expand...
April 2010.
 
Yep, I remember the conversations, @dividedbyzero... But stating the fact that there is a massive amount of USB tech out there in consumer hands right now is just that: stating a fact. I'm not putting LightPeak down, I'm saying it's starting out well behind the curve, and has an uphill battle to "replace USB" as that Guest was stating.

And yes, I too saw the demo prototype connections using USB. I've been hoping that combo LightPeak/USB connector was going to be the standard. Has there been actual confirmation that this first generation of LightPeak will truly use those USB 3.0 connection ports, or have the connectivity options that were referenced at launch? I'm hoping so, would like to see LightPeak get every advantage to guarantee success.
 
matrix86 said:
This is pretty much floppy/cd drives all over again. Floppy was the way to go, then cd came in. For the longest time, machines had drives for both, and we are now seeing the final end of the floppy. This is what's going to happen with USB/LightPeak. Both will be integrated onto the same machine until USB fades out (which like the floppy, will take several years). Face it, Light Peak is the future. It's so much better than USB and has much more potential. It's just going to take some time to fully integrate. This will not be an "overnight" thing like people seem to think. Stop trying to hinder the future and look at the past...it will show you how the future's going to go.
IEEE 1394 is still trying to outdo USB, but lightpeak will come right out way ahead of USB which gives it the advantage. As USB improves, so will light peak until USB realizes it can't keep up. Poor IEEE 1394 just needs to give up, lol.
Click to expand...

I am in no way trying to hinder the future. Check my other posts, I am in agreement that LightPeak has benefits that far outweigh what USB can provide. I love the fact that it has so many protocols that can be driven from a single connection, the possibilities are actually staggering - if you could see the rat's nest of cables hidden in my entertainment center behind my HTPC, you'd know why I love the concept! :)

You should also note that you are just reiterating exactly what I keep telling people who are saying LightPeak should just "replace USB." It will be a process, because there is a tremendous amount of USB equipment in the marketplace now. Many just seem to want to snap fingers and make the switch, without considering the actual cost that switch can incur for consumers, in the short term. Having both options, with LightPeak being so superior that consumers can't help but choose that format for any future purchases of equipment, is the best way to facilitate LightPeak truly replacing USB.
 
You can think of it from the point of view that, lots of people have USB devices etc.... But then I can remember when so, so many people had serial and parallel devices and would not give them up.

USB will die like serial and parallel devices. Slowly possibly, but long after the manufacturers have stopped making the devices, motherboard makers will still be including a single USB 3.0 backwards compatible port, so a certain segment, who might even still own a 24pin dot-matrix parallel printer, can continue using their Iomega 250MB 3.5 inch portable USB HDD for backing up critical data, in 2015.

I betcha some of them still own an Atari Jaguar.
 
gwailo247 said:
I'm not Intel bashing, you're right, they may not have explicitly advertised it. I guess I just made an *** out of me by assuming that a game changing chipset is going to have support for a graphics standard that has been around for a few years. I know it doesn't make a difference for the average notebook or tablet consumer which is going to be using this chipset and will not need DX11 graphics, but I still feel that this is something that would hold me back from purchasing this chipset for my own needs.

I'm never an absolute fanboy of everything, I laud what works, I criticize things that I feel have problems. I feel that lately, starting with USB 3.0, now the previously mentioned issues, Intel has been focusing more on the marketplace than the product. And I don't like it when companies do that. So I'm going to ***** about it. That is not going to stop me from being a loyal Intel customer. =D
Click to expand...

Well obviously it would have been nice if they'd included it. I'm simply sick of posts like the one I also quoted above from AMD fanboys trying to bash Intel. Sorry if I came off as accusing you of lying.
 
Do me eyes see two princetons? <img src="http://img203.imageshack.us/img203/2141/51499282.png" alt=""/>
 
princeton said:
lawfer said:
Do me eyes see two princetons? <img src="http://img203.imageshack.us/img203/2141/51499282.png" alt=""/>
Click to expand...

Broken link is broken :(

Anyway I pasted it. What do you mean by "two"
Click to expand...

Hahaha, it's just that there are two of your comments; one with your username starting with a capital P, and the other with a lower-case p.

I'm kind of a grammar nazi, so even that little difference made me doubt your authenticity. ;)
 
You must log in or register to reply here.

Similar threads

Jimmy2x
Fiber optic researchers showcase speeds 4.5 million times faster than average home broadband
Replies
18
Views
2K
Levi1922
L
midian182
Pretty uncool: Half of tested watercooler radiators fail to use advertised copper
Replies
5
Views
171
Endymio
Endymio
Shawn Knight
Bitcoin surpasses $69K to hit new record, eclipsing 2021 peak and overcoming 2022 crypto...
Replies
16
Views
257
Crabjuice
C

Latest posts

Back