Intel's 50Gbps Light Peak Successor

Barence writes "Intel has unveiled yet another high-speed optical interface – before its long-awaited Light Peak connector has even reached the market. The Light Peak optical interconnect can transfer data at 10Gbps in both directions, and is touted as an all-in-one replacement for USB, DisplayPort, and HDMI. The new interface uses an indium phosphide hybrid laser inside the controller chip — a process that Intel calls silicon photonics — rather than using a separate optical module, as with Light Peak. And by encoding data at 12.5Gbits/sec across four laser beams of differing wavelengths, the connector yields a total bandwidth of 50Gbps, five times that offered by Light Peak. 'This is not a technology that's ten years away, but maybe three to five years,' Intel fellow Mario Paniccia announced. 'Light Peak, as we've stated, will launch next year.'" HotHardware quotes Intel in more detail on the difference between the two programs: "This research is separate from Intel's Light Peak technology... Light Peak is an effort to bring a multi-protocol 10Gbps optical connection to Intel client platforms for nearer-term applications. Silicon Photonics research aims to use silicon integration to bring dramatic cost reductions, reach tera-scale data rates, and bring optical communications to an even broader set of high-volume applications."

Comments

50 Gbps ought to be enough for anyone

[maharg]

.. now could you just roll that out globally please :-D

Re:50 Gbps ought to be enough for anyone

[imamac]

I see no reason why LightPeak products can't be sold internationally...

Just what gamers need

[Locke2005]

How long until an optical mouse using this technology is available?

Re:Just what gamers need

[Anonymous Coward]

I wish hard-core gamers would stop thinking of themselves as gods or something. A human being can't even fill the 1.5 Mbps bandwidth offered by USB 1.1 so let's not talk about this for a damn optical mouse.

Re:Just what gamers need

[Khyber]

"A human being can't even fill the 1.5 Mbps bandwidth offered by USB 1.1"

Bullshit, we make web cameras that can do that now.

Re:Just what gamers need

[c6gunner]

Yea, I guess with the average persons diet being what it is, we're getting closer and closer to being able to fill that much bandwidth.

Re:Just what gamers need

[Anonymous Coward]

"A human being can't even fill the 1.5 Mbps bandwidth offered by USB 1.1" [with a mouse]
context.

Re:Just what gamers need

[Khyber]

Could still be done with a mouse. Just make the controller output at maximum bandwidth. All it takes is one shitty driver.

Re:Just what gamers need

[AHuxley]

$200 from Razer, real Tron like this time :)

Re:Just what gamers need

[Locke2005]

I wish hard-core Anonymous Cowards would stop thinking of themselves as geniuses or something and learn to recognize humor when they see it. Of course 50 Gbps is overkill for a mouse, that's why it's funny!

Re:Just what gamers need

[supertrinko]

Your parent never said he thought of himself as a hardcore gamer, or showed in any way that he thought of himself as a god. I don't know about you, but USB 1.1 is not enough for many tasks.

Re:Just what gamers need

[Anonymous Coward]

In addition to USB cameras; USB Audio devices, USB video display adapters, high-precision USB graphic tablets, USB haptic input/feedback devices, etc.. Personally many of my own USB projects easily cross that threshold. It's not difficult when you look at taking large bursts of data in time slices less than 1 second, not to mention packet overhead. > ~100 to ~150k in 100ms? Absolutely, 100ms is an eternity.

Why optical?

[Peach Rings]

USB and HDMI cables have to be really short anyway, isn't optical overkill? I mean, you have copper on both ends, having an ultra-high-bandwidth hybrid laser in the middle isn't going to perform any miracles. Just run parallel wires instead of serializing everything and you have all the throughput anyone could possibly use.

Re:Why optical?

[Rakishi]

Just run parallel wires instead of serializing everything and you have all the throughput anyone could possibly use.

Why would you want the link to be slower? Hint: There's a reason everything is serial now rather than parallel.

Re:Why optical?

[Locke2005]

Just run parallel wires instead of serializing everything and you have all the throughput anyone could possibly use. Too bad the people that designed SATA didn't think of that!

Re:Why optical?

[klui]

Running too high a clock on an electrical parallel interface causes discrepancies in trace length to be an issue so it's simpler to use a serial interface. In addition, interference between different wires may make the connection unreliable in a parallel interface.

Re:Why optical?

[radicalpi]

I vaguely recall some wide, flat cables. We used to use them for connecting optical drives and hard drives. What were those things called again? Ah, I forget...

Re:Why optical?

[MobyDisk]

To explain the point the parent is alluding to: When you run copper wires at high bandwidth it induces a magnetic field. The magnetic field then induces a current in the neighboring wires. This is crosstalk. The more wires you have closer together, the more crosstalk. This is part of why everything is moving from massive parallelism (ribbon cables) to high-speed differential signaling. You use only two wires, and the two wires always send the opposite signal. When one wire sends a 1, the other sends a zero (that's a simplification). And vice-versa. Optical cables don't experience crosstalk.

The other major reason for the shift is that ribbon cables get expensive, and are a pain to route.

Examples of things that use this:
- USB
- SATA
- DVI, HDMI
- Ethernet

Re:Why optical?

[Deflatamouse!]

There is a reason that the industry have been trending towards serial and away from parallel buses.

It's been a while since I've done an transmission line and bus design work. Let me see if I can explain this in 'lay' terms:

To implement a parallel bus, you have to have each and every wire be within a certain variance. Your driving and receiving chips also need to be able to send and receive the data within a certain variance. This is because you typically send your data, say a 32-bit word over a 32 wire bus, across the bus at the same time. If the wires (and drivers and receivers) do not match up, your data will be scrambled on the other end of the bus.

The larger your chips (because you need all the drivers and receivers to send the parallel signals) or the more wires you have, the variance between the parts becomes harder and harder to control because of manufacturing limits. The trick is to design your entire system to tolerate the variances of each individual parts so that they will still work together.

But at the same time, you want to increase the speed of the bus (because having 20,000 wires is just not so practical). This is a force in conflict with what you're trying to achieve because an increase in speed translates to less tolerance in the system for parts variance.

At some point between increasing parallelness and higher and higher speed, the increase in variance will exceed the system's tolerance, and the parallel bus becomes impossible to implement or unreliable.

This is why bus designers have been trending towards serial interfaces, because that at least takes most of these variances out of the equation (it's still there but less influential).

The other trend is clock encoding. Instead of sending bits synchronously, or sending a strobe (a separate clock) signal along with the data. Now we 'encode' the clock into the data, using encoding such as the 8B/10B encoding. The receiving circuit can then 'retrieve' the clock from the data signal (it basically allow you to identify each set of data from each clock cycle, and detect problems). Serial interfaces are also usually accompanied by training sequences at start up (may be software implemented) to adjust various parameters to make the data transmission ideal for the environment.

Re:Why optical?

[bell.colin]

What?? HDMI cables do not have to be short, I use a $90 HDMI over Cat-5e converter box over 70-feet and run 1080p @ 120hz to my 60" LCD with no problems. And that's with (1) 10' HDMI cables, (2) 1.5' Cat-6 into (2) Cat-5e jacks with RJ-45 crimped on the other end to the converter with another 10' HDMI cable.

Really short??? And USB is way lower bandwidth than HDMI.

Re:Why optical?

[BusterB]

Also, note how this is not a single serial 50 Gbps link - it's 4 parallel 12.5 Gbps links. You can run light in parallel with no interference, the trick is to make sure that each independent channel uses a different wavelength instead. So, they are doing it in parallel. Some 100 Gbps ethernet standards use 10 parallel 10Gbps lasers running at different wavelengths, but they are amazingly expensive because of this.

Re:Why optical?

[electrostatic]

"...use 10 parallel 10Gbps lasers running at different wavelengths, but they are amazingly expensive.."

From TFA

And by encoding data at 12.5Gbits/sec across four laser beams of differing wavelengths, the connector yields a total bandwidth of 50Gbits/sec, five times that offered by Light Peak.

Re:Why optical?

[Khyber]

"Optical cables don't experience crosstalk."

I wouldn't be so sure of that. We've seen the Emerson effect in plants, and looking at it on a quantum level it seems similar effects might apply to optical cabling.

Re:Why optical?

[Anonymous Coward]

In a sense, modern Ethernet cables do this too. Gig-E uses all four wire pairs of a UTP cable and requires that each pair have a slightly different rate of twist along the length of the cable. The varied rate of twist reduces crosstalk between wire pairs.

Re:Why optical?

[dbraden]

I also remember reading somewhere that it's easier to achieve high speed over a serial interface because once you start dealing with very high speeds the timing differences of when a signal arrives at the destination become a big factor.

In a super simple, flawed I'm sure, example, assume you have an 8-bit interface (and ignore other lines required). When you send a byte down the line, you have each bit traveling down it's own data line. When they reach the other end, you have to wait until you have all eight before you can reconstruct the original byte and hand it off.

To human perception, those 8 bits will arrive simultaneously, but in computer-time (bullet-time?) it can seem like ages waiting for all 8 slots to fill, and will probably become more and more out of sync as time goes by, forcing all 8 lanes to shutdown periodically (maybe one of the data lines has a kink, flaw, or is simply slightly longer than the others).

Right or wrong, that's my understanding of it ;)

Re:Why optical?

[maxume]

From the comment you replied to

So, they are doing it in parallel.

He is pointing out that current products that do this are expensive, not arguing that this new thing will be, or anything like that.

Re:Why optical?

[ooshna]

So to have longer hdmi cables you just have to shell out $90 for a converter box plus the cost of an extra hdmi cable and the cat-6 cable. That's not too much of a waste or pain in the ass (insert trademarked sarcasm punctuation here)

Re:Why optical?

[Anonymous Coward]

You are correct. I think Light Peak is simply 2-lane PCIe and a bunch of marketing BS. USB3 is 1-lane and uses cheap copper cables, and its 5Gbps is plenty fast for hard drive transfers.

Re:Why optical?

[QuantumRiff]

I would love this for ceiling mounted projectors or wall mounted TV's.. you could move the equipment far away from the TV, such as in a closet behind the couch..

Re:Why optical?

[fjanss]

USB and HDMI cables have to be really short anyway, isn't optical overkill?

It is a replacement that, because it is optical, does not need to be limited by "really short cables". If the technology is cheap enough, I would love to have webcams at 100m distance instead of expensive ethernet cameras (as an example).

Re:Why optical?

[Kizeh]

Well, kinda. If you have physically separate fibers, it's not much of an issue. If you have different wavelengths on the same fiber, they unfortunately do interact, for example http://en.wikipedia.org/wiki/Four-wave_mixing, but there are other non-linear phenomena as well.

Re:Why optical?

[Anonymous Coward]

USB and HDMI cables have to be short because the copper won't support longer distances. If you only imagine them used for a desktop computer, there's not very far for the cables to go. But what if you want your computer used for a large home theater? You want the CPU in another room (possibly on a different floor) because of the fan noise, you want the HDMI cable to go to the projector in the back of the room, and you want the USB IR receiver at the front of the room. Since the cables are going to be routed through conduits in the walls, you could easily need 50-100 feet of cable.

What about a digital signage application? In airports they might have 100 monitors distributed all around a concourse, all showing the same arrival and departure information. Due to the signal length limitations on HDMI, you need dozens of PCs scattered about the airport to support all those monitors. It would be a lot easier to have a single computer in a central location distributing the display to all the monitors.

dom

Re:Why optical?

[elashish14]

Less power. Optical interconnects can operate at much lower power than their semiconductor counterparts. Other benefits include potentially using multiplexing (sending additional information along the same cable, differentiated by some property of the signal such as the wavelength of the light used) to enhance the signal bandwidth. And higher-frequency switching: light could, in theory, be modulated in ~10^-15 seconds, the optical frequency whereas electronic frequency is hard to get smaller ~10^-9 seconds (my figures may be off). Of course, no one has actually implemented anything practical at these higher switching frequencies yet, so who knows if this will pan out.

Re:Why optical?

[Alex Belits]

This is the ONLY real problem with parallel interfaces -- crosstalk is a complete red herring because no one in his right mind will approve a cable, serial or otherwise, that will mess up data on another cable laid in parallel. Parallel interfaces of the past rely on single clock for all lines, so they can fill bus-wide buffer in one cycle.

However with multiple lines, each with its own synchronization and with a larger buffer on the receiving end, clock skew is merely latency -- you have to wait for every bit to complete its cycle before you can push the received data word to the bus. So parallel interfaces are possible, they just require different mechanism of data transfer and synchronization. It is more expensive, but if you really need this speed, it is easily achievable.

Re:Why optical?

[FiloEleven]

Is the Emerson effect due to something like crosstalk? I just looked it up because I was intrigued, and WP and what it links to are of little help, but it looks to me like the two frequencies activate different systems within the plant cell.

Re:Why optical?

[dargaud]

This is part of why everything is moving from massive parallelism (ribbon cables) to high-speed differential signaling.

Another reason is keeping the signals on the various wires synchronized. At the frequencies we now use, say 50Gbps, that's only 6mm between one bit and the other. If one wire of the same cable is only 1mm longer than the other you run a one in 6 chance of mistaking the bit with the previous or next one. Adding delay lines add lots of complexity.

Re:Why optical?

[Anonymous Coward]

He was talking about 100Gbps Ethernet, not this Light Peak watchamagig. It helps on slashdot if you know how to read.

Re:Why optical?

[bigdaisy]

Let me see if I can explain this in 'lay' terms:

To implement a parallel bus, you have to have each and every wire be within a certain variance.

Sorry, you lost me at the end of the very first sentence. I realise now why you put the word "lay" in quotes. What is a "variance"?

Re:Why optical?

[kestasjk]

The other trend is clock encoding. Instead of sending bits synchronously, or sending a strobe (a separate clock) signal along with the data. Now we 'encode' the clock into the data, using encoding such as the 8B/10B encoding. The receiving circuit can then 'retrieve' the clock from the data signal (it basically allow you to identify each set of data from each clock cycle, and detect problems). Serial interfaces are also usually accompanied by training sequences at start up (may be software implemented) to adjust various parameters to make the data transmission ideal for the environment.

Isn't Manchester encoding / differential Manchester encoding (having both clock and data going down one line) really old?

Re:Why optical?

[gmarsh]

Replace "variance" with "length". Every line on a high speed parallel bus on a circuit board has to be almost exactly the same length, otherwise the delay you get in the longer lines can make the bits on those lines arrive too late. Likewise, valid data can arrive too early on a couple of lines if they're too short.

If you look at the memory bus on a modern motherboard, you'll see lots of "squiggly" traces, traces which loop back on themselves, etc. This is done to make the short lines longer so that the length of all the lines match up.

About the only "massively parallel" bus that you'll still find on a computer is the SDRAM bus, everything else (SATA, USB, PCI Express, HyperTransport, etc) is all serialized. But I fully expect "parallel" DRAM technology to hit a wall soon, and DDRx memory to be replaced with a something that uses multiple high-speed serial interfaces instead of single wide parallel bus. I'm also expecting Rambus to rear its ugly head again, when that happens..

Re:Why optical?

[Khyber]

The Emerson Effect is optical crosstalk that occurs when 660-670nm light is mixed with 720-740nm IR. It ends up enhancing photosynthesis, stimulating the same phytochemical systems and increasing the rate of photosynthesis.

Re:Why optical?

[MikeBabcock]

USB and HDMI have to be short precisely because they're wire interfaces. With optical, we'd be able to have HDMI connections hundreds of feet long presumably.

There are already HDMIopticalHDMI connectors available, much like the old parallel cable extenders, but having the optical protocol in place would eliminate the extra adapters.

PS you don't understand why SATA replaced PATA or SAS replaced LVD SCSI either, do you?

Re:Why optical?

[MikeBabcock]

Don't forget the lack of heat in the cable.

Electronic cables have to be able to withstand the heat generated by the electrical resistance in the wires. Optical cables have almost no such problem.

Re:Why optical?

[adisakp]

There is a reason that the industry have been trending towards serial and away from parallel buses.

It's still possible to combine multiple high-speed serial buses in parallel as long as you handle skew independently per transmission line. That is what happens with PCIe - you can have x1, x4, x8, x16 lanes where effectively the x16 lanes are just 16 x1 lanes in parallel.

Re:Why optical?

[Anonymous Coward]

Mod parent FUNNY, not 'Interesting'...

Wow, some people really forget there history quick.

Re:Why optical?

[Anonymous Coward]

Yes - on old school parallel ports for example, the recieving device would only signal 'byte recieved' once all bits are present and you could then move on to the next byte.

It's also why PCI Express is actually several, serial, but independent, lanes. A PCI-Express x4 slot, for example, is really four independent lanes each of which run with their own clock synchronisation, so lane 1 and 2 could easily be 'out of phase' from each other (it's not parallel, but side-by-side serial).

Re:Why optical?

[Unequivocal]

IIRC, parallel cables had much shorter max lengths (SCSI, parallel printers) b/c of noise at longer lengths? Am I misremembering?

Re:Why optical?

[Deflatamouse!]

This is true, and I think the main point is that such a parallel bus is made possible with the new serial technology. You will still need to sync up each bus, but it can be done at a higher granular level than individual bits as it was done in traditional parallel buses. For example, you can transmit packets of data through the serial bus, containing information to help the receiving end piece together the larger block of data. This isn't possible with single wires since a single bit can't hold this type of information ;)

HDTV Warranty

[HTH NE1]

The way things are going, extended HDTV and HD monitor warranties are going to need interface obsolescence coverage.

Re:all this rhetoric about what we've done is fine

[Anonymous Coward]

Don't forget to take your meds tonight.

Obligatory...

[sznupi]

Imagine a beowulf cluster using those!

Re:Obligatory...

[Unequivocal]

Keeping the dream alive!

Fast Disks?

[PineHall]

Is there a storage device today that can deliever 50Gbps speeds? Right now I don't know of any, but that does not mean there will not be some in the future. It seems this technology is getting way ahead of everything else.

Re:Fast Disks?

[radicalpi]

Exactly, who needs a 50Gbps connection when they have a drive that maxes out at a couple hundred Mbps.

Re:Fast Disks?

[magarity]

Is there a storage device today that can deliever 50Gbps speeds?
 
Yes, they're called enterprise grade SANs. A good one is faster internally than the latest fiber connection and just begging for an upgrade to this new tech.

Re:Fast Disks?

[Anonymous Coward]

Exactly, who needs a 50Gbps connection when they have a drive that maxes out at a couple hundred Mbps.

Think compression, video compression for example.

Re:Fast Disks?

[Takahashi]

Its naive to think that buses are only good for storage.
1080p video is 1080*1920pixels/frame*32bits/pixel*60frames/second is roughly 4Gb (uncompressed). If history is any indication displays will get larger and denser (also remember bandwidth needed will scale by the square of the number of lines). Or what if you want an external GPU? 16x pci express is about 32Gbits/sec. Also more speed reduces latency, which may be helpful too depending on what you use it for.

Re:Fast Disks?

[Mathness]

Why don't they use transmission arrays (TA)? It seems obvious to me that SAN plus TA would have a massive packet delivery boost. Of course the downside is when something bad happens, you tend to end up with a lump of coal. ;P

Re:Fast Disks?

[asc99c]

External graphics cards? These have been tried before for laptops, but electricity only moves down copper traces at a very limited speed. This sort of connection could decrease the latency as well as bandwidth. Potentially other external components could be done the same way - e.g. additional memory in a laptop base station. The bandwidth is enough to support it, and the latency should be good too.

Cause copper is too slow

[Sycraft-fu]

Best bandwidth we see out of something like DisplayPort now is about 17gbps. That works fine for today's displays, but we'll need more if we want better. Ideally we'd like to move to more bits per pixel so that we can have a large colour gamut without banding and perhaps HDR displays, we'd like a lot higher rez so you can't see individual pixels, and 120Hz (maybe more) would be good so motion is dead fluid and maybe for 3D. That is going to need a shitload more bandwidth.

Unfortunately we seem to be running in to limits on what copper can handle. Note that we are already doing parallel communication. DP is 4 parallel lanes to get 17gbps. More lanes = more cost and more wire. Trying to go massively parallel would be a problem.

Moving to fiber may be what is needed these days. We seem to be butting up against the limits of what we can easily and cheaply get copper to scale to. I don't know about you, but I'd love to see a universal bus. Drives, mice, displays, everything all run off the same bus. Would nicely simplify things. However to do that, it has to have some killer bandwidth (10gbps is ok for now, but not for long) and it has to be cost effective.

Re:Cause copper is too slow

[Surt]

You really want 240hz for the combination of fluidity and 3d at the same time, particularly for games.

Re:Cause copper is too slow

[Peach Rings]

I just had a really cool idea.. what if a part of the bus were directly tied into the CPU bus? So you could connect two devices and machine A could be the master and it would have direct access to other CPU. Some time-expensive operation? Make B do it, and set up the flags so that B's cpu looks into A's memory to do memory operations.

What's most exciting is when you think of low-power devices like phones and ipads. Have a computationally-intensive task for your device? Walk up to your desktop computer, or a coffee shop lunch-box-sized booster server box, plug in the cable, and your device is instantly 10,000 times faster. No need to install the software performing this difficult task onto the server, no need to transfer data files or results. It simply makes your device super fast, and you interact with it normally.

Just an idea

Re:Cause copper is too slow

[MikeBabcock]

To be honest, I'd rather have a display interface that understood an arbitrary N channels of video, much like current audio standards (dts, DD, etc).

A game could feed two screens worth of data down a high bandwidth HDMI cable at 60Hz for use as a 3D video game and the receiver could either be a TV that flipped back and forth between the feeds at 120 or 240Hz, or it could be a pair of projectors with an HDMI pass-through port each projecting one eye's display simultaneously for use with polarized lenses.

The same HDMI cable could be fed three screens worth of data for a wider field of view, or two screens worth for a multiplayer game, leaving it up to the TV to render side by side or above and below, or picture-in-picture.

Future-proof extensible specifications would be nice. *sigh*

Re:Cause copper is too slow

[Unequivocal]

From the best as I understand, the HS cross-connect to the faster machine is not the hard part of making a system that works like you describe. OS and hardware interactions are (protecting ram, etc) are the tough bits. Thinking machine went after an industrial version of this, and there are lots more. Oh - even beowolf! Wow, I used beowolf in a slashdot post.

Parallel?

[PIPBoy3000]

I look forward to using parallel ports again. Of course, I may have problems figuring out which of the million pins just got a tiny bend in them.

Clue injection

[Anonymous Coward]

USB and HDMI cables have to be really short anyway, isn't optical overkill? I mean, you have copper on both ends, having an ultra-high-bandwidth hybrid laser in the middle isn't going to perform any miracles. Just run parallel wires instead of serializing everything and you have all the throughput anyone could possibly use.

Wow. Where to begin.

Seen the price of long HDMI cables recently? Go check that out and come back with a clue. No, that isn't all just markup gouging or whatever. The reason for that cost is that high frequency signals on copper require several expensive techniques to overcome signal loss. Fiber doesn't have that problem. Optical signal cables don't have to be short. They can be miles long.

Introducing parallel signal lines would only multiply the cost of signal cables due to even greater signal problems created by crosstalk.

Serial has replaced parallel at every level of communication, right down to motherboard traces (PCI-Express, QPI, etc.) because high frequency serial is MUCH easier (read cheaper, more reliable, etc.) than parallel. The only designs considered today for new buses are serial and trunked serial.

Finally, photonics is real today. I'll repeat that; PHOTONICS IS REAL. The problems have been solved. Fabs make lasers on chips now. They are integrated with the silicon circuits. They are cheap. They are fast. They cost nothing. Try to keep up.

Re:Clue injection

[Daetrin]

"Finally, photonics is real today. I'll repeat that; PHOTONICS IS REAL. The problems have been solved. Fabs make lasers on chips now. They are integrated with the silicon circuits. They are cheap. They are fast. They cost nothing. Try to keep up."

So here's a question, just because you didn't include "reliable" in your list of superlatives. What's the mean time to failure for these laser on a chip devices? Years or decades or longer? I'm still using a 26" (i think) SD tube TV at home that i bought back in... 2000 i think. I _might_ upgrade to a HD LCD TV this year. Or maybe next year. I'm not the type to be constantly upgrading to the latest and greatest (okay, except for game consoles, but those have a built in delay period) and i certainly don't want to get cables that are going to die before the device they're plugged in to. If they're guaranteed to last at least a decade or two that's _probably_ good enough for me.

Re:Clue injection

[Khyber]

"What's the mean time to failure for these laser on a chip devices?"

For the typical laser diode, with todays construction methods, minimum 100,000 hours. The rest of the stuff attached to the laser, I'm not so sure.

Re:Clue injection

[PitaBred]

...I can get a 50' HDMI cable for $30 US. How is that expensive? You pay any more and it most certainly is markup and gouging

http://www.monoprice.com/products/product.asp?c_id=102&cp_id=10240&cs_id=1024005&p_id=2110&seq=1&format=2

10Gbps??

[thewils]

Not sure what that is...can I get that converted to mp3s/sec?

Re:10Gbps??

[wjh31]

About 400-250mp3/s at 3-5MB/mp3

Re:10Gbps??

[Anonymous Coward]

2560 mp3s/sec

Re:10Gbps??

[DeadDecoy]

Or $40,000,000/s by RIAA standards.

Re:10Gbps??

[Fumbili]

MP3 sizes are too variable. You need a fixed metric that the common man will understand like...Library of Congresses/sec

Re:10Gbps??

[aristofeles]

So lets make it more clear: according to gizmodo, " Now imagine transferring the entire printed catalog of the Library of Congress in a minute and a half. "

Re:10Gbps??

[pinkushun]

For near studio quality of 192 kbps (kilo bits) audio, 8 bits in a byte = 24 Kb (kilo bytes)

1 GB = 1,000 MB
10 GB = 10,000 MB
10,000 MB = 10,000,000 Kb (kilobytes)

(10,000,000 / 24) = 416666 seconds = 6944 minutes, which gives us...

115 hours of good quality audio. That's 4 days of solid music every second.

Re:10Gbps??

[bdforbes]

Okay how about in football stadiums per superbowl? It's the only thing I understand!

I still wonder...

[fuzzyfuzzyfungus]

I still want to know what Intel has up their sleeves for this "light peak": You can get 10Gb optical interfaces right now, off the shelf; but they are quite expensive, and the connectors and cabling aren't something you'd trust a noob to get too many mating cycles out of. What are they going to do to change that? Are there some substantial economies to be had if you compromise on max link length? Do they have some clever new optical connector design?

A 10Gb/s SFP+ optical interface is ~$200(bracketed, I'm sure, by Ebay and Cisco, on the low and high ends, respectively). That is not exactly going to fly in consumer electronics land(Oh, double the BOM cost, no problem!). Therefore, I can only conclude that Intel has some clever plan. If they have a clever plan, though, why are they talking about consumer electronics, rather than absolutely cleaning up in the relatively short range, high-speed, datacenter interconnect market?(Not that they'd necessarily ignore the consumer market; but if they can do what "light peak" promises, one would conclude that they can do 10Gb ethernet, at least over modest distances, substantially more cheaply than anybody else. That would be worth a bundle.)

Re:I still wonder...

[Unequivocal]

Presumably they figured out how to fab this on a chip which is the magic. And I'd guess they'll clean up all over with it..

Don't think in terms of old interfaces

[Just_Say_Duhhh]

Sure, TFA says Light Peak is a replacement for USB, HDMI, etc., but you don't have to think in terms of those soon-to-be-obsolete implementations.

Using optical instead of electronic signaling in high-speed interfaces has great advantages. You said it yourself - USB and HDMI cables have to be really short. When you have electrons racing down the wire at high speeds, they tend to crash into the end of the cable with a BANG! (causing reflections, RFI and lots of nasty problems). Optical signals can be boosted enough so that longer run-lengths are possible, without the problems found in electronic signaling.

Parallel? In electronics, we've moved past that. When you only ADD lines, but MULTIPLY your problems as you increase speed, parallel signaling loses its luster.

If we solve the issue of optical connections (it has to be cheap, reliable AND fast), we could even implement parallel optical pathways, and data rates would skyrocket!

Delicious DRM.

[Alex Belits]

I just have one question -- is the purpose of all this crap to make a device that only few manufacturers can produce, then make sure that only DRM'ed to Hell version is available on the market?

HDMI DRM is for all practical purpose defeated (YA, RLY) by the use of mass-produced $100-$300 HDCP strippers in homemade DVRs -- now our beloved content providers want hardware companies to build something else, easier to keep out of consumers' hands?

Re:Delicious DRM.

[westlake]

HDMI DRM is for all practical purpose defeated by the use of mass-produced $100-$300 HDCP strippers in homemade DVRs

Mass-produced by who and at what risk?

The homemade DVR is a pleasant - if geekish - project.

But the father of three kids rents - or more likely buys - the video from Pixar - and that is where the money is.

Re:Delicious DRM.

[Alex Belits]

Mass-produced by who and at what risk?

http://www.hdfury.com/

The homemade DVR is a pleasant - if geekish - project.

And that covers absolutely everyone who will ever bother to copy DRM'ed video. Demise of NTSC and Comcast's eagerness to DRM each and every channel will eventually force everyone who has a homemade box to add such a device, and then HDCP will be about as effective as DVD CSS.

But the father of three kids rents - or more likely buys - the video from Pixar - and that is where the money is.

And he would buy it even if DRM is broken. VHS tapes and CD were sold and rented just fine without DRM, the problem is that content producers want to squeeze more and more profit from the same content.

thats like 100 porno videos simultaenously!

[NynexNinja]

wow thats alot of porn

It's all part of the Connector Conspiracy

[Ancient_Hacker]

It's all part of the Connector Conspiracy.

I must have $500 tied up now in 40/80 pin IDE cables, SATA cables, 8-Bit Apple SCSI, four or five other flavors of SCSI interconnects-- mini- sub-mini, regular and LVDS, VGA cables, HDMI cables, USB type A, B, and Mini. Let's not forget the big bag of "RCA Phono" cables, to and from eighth-inch mono and stereo plugs. Then all those offbeat motherboard to PCI-slot Parallel port flat cables. ANd parallel-port printer cables, and who could forget serial cables, DB9, DB25, gender-changers, and breakout boxes. And the various internal flat- SCSI cables and connectors. And the various Vidio connectors on iMacs-- at least four varieties there. Somehow, no matter how many bulging cardboard boxes of cables and adaptors I have, each month I have to make a new trip to BEstBuy to purchase some overpriced new cable. I thought things would plateau for a while with the cheap SATA cables, but noooooo, we better start saving up for a whole new series of optical interconnects.

Re:It's all part of the Connector Conspiracy

[jackchance]

if you are shopping for cables at BestBuy you have serious problems.

check out monoprice.com

Re:It's all part of the Connector Conspiracy

[bill_mcgonigle]

It's all part of the Connector Conspiracy.

You're complaining about a plethora of cables on an article about a one-cable-to-rule-them-all technology, and calling 'conspiracy'?

I too have a closet full of obsolete cables and would like to put that dark-alley of technology behind me.

At last

[sqrt(2)]

Perhaps my dream of having 1 port for everything, peripherals, storage, display, power even, will be achieved. Just a line of identical ports on the side/back of the computer.

Re:At last

[mlts]

I'd like to see a connector with both an optical connector (for two way communication) and a copper connection for power.

I worry though. Give Joe Sixpack a single mode fiber optic connector with a warning "do not look down connector with remaining eye", and he probably will need Braille to do his next computer stuff.

Re:At last

[SteeldrivingJon]

"I worry though. Give Joe Sixpack a single mode fiber optic connector with a warning "do not look down connector with remaining eye", and he probably will need Braille to do his next computer stuff."

TOSLINK optical sockets have little cover flaps. I suppose they could figure out a way to have such auto-retracting flaps on both socket and cable.

Um, in fact, it's starting to sound downright anatomical.

Re:At last

[acnicklas]

I thought lasers such as TOSLINK were eye-safe? Don't have a reference, but I'm pretty sure they're only a Class I laser.

Re:At last

[mlts]

Makes sense. Anatomical or not, the flaps will mean that Joe Sixpack has to defeat some type of safety system in order to burn his eyes out. However, Joe Sixpack Jr. will likely bypass it, and the maker will end up on the short end of a lawsuit because of the kid.

Perhaps a system that verifies the connection over the copper first, then starts the optical communication. This way, it will take some doing for Joe Sixpack or Aunt Tillie to be able to eyeball the laser diode.

Re:At last

[fuzzyfuzzyfungus]

Unless you have some particularly classy TOSLINK gear, there are no lasers involved, just LEDs(and at 650nm, so all the eye's brightness adjustment and blink features work normally).

There might have been some good reason for it back in the early 80's; but the fact that TOSLINK is optical at all looks like crazy overkill by today's standards. S/PDIF signals can be transmitted electrically over a single RCA cable just fine at modest distances(3.1Mb/s just isn't that challenging), and the cost constraints on TOSLINK(LEDs rather than lasers, usually just plastic fiber), mean that its range is scarcely any better.

Re:At last

[phyrexianshaw.ca]

The goal was to prevent digital noise on the line, not to offer anything copper didn't already have

Toslink allowed people to bridge from the DVD player to the receiver, across the AMP, and not pick up stray noise. beyond that, it's only still around because people think it's cool to say: "yeah, I needed an optical cable to hook up to my receiver."

Re:At last

[owlstead]

Ehm, ONE copper connection? Give Joe Sixpack a wire that actually has to run back to the computer and he will certainly hang himself with it, never mind that eye. Or do you want to connect all devices to the radiator of the central heating system?

Re:At last

[aXis100]

Yeah, but even that is barely an excuse. It's not that hard to keep noise in a digital inteface below the logic threshold.

Marketing Problem

[ISoldat53]

Did Intel commit a marketing mistake by announcing a follow on product before they come to market with the first product?

Re:Marketing Problem

[maxume]

No.

Re:Marketing Problem

[JSBiff]

That line of reasoning seems to be a little self-limiting, particularly when you are talking about advances from a single company. What I mean is, although sometimes you might decide to 'skip a generation' (I think perhaps in this case, maybe Intel should skip Lightpeak and move straight to the faster tech from the get-go), in general, no one can afford to never have any revenue generating products on the market because they are always just 2 or 3 months from releasing the next product. That's a fast way to go bankrupt. At some point, you have to just sell a product for awhile, perhaps even intentionally delay the next generation of a tech if it's ready 'too soon', so that you can recoup your costs from developing the previous generation.

There's also the issue of standardization - 'standards' that are obselete in three months, were never really standards. For a common interconnect that can be used with TVs, monitors, home theater systems, video game systems, recording devices (TiVo, etc), Camcorders, etc, stability/longevity of the standard is just as important as technical specs - after all, it doesn't do you any good that there's a technically superior cabling system if all of your current devices use the old cables and connectors. Sure, people are willing, eventually, to replace things, but they don't want to do it within a year of the previous standard being adopted. Honestly, I think it's still way too soon to be replacing HDMI.

Re:Marketing Problem

[Anonymous Coward]

No, not at all.

Oblig XKCD

Re:Marketing Problem

[Shotgun]

I still think it is to early to replace DVDs, yet we see this insane push for Blu-ray, and now there are other technologies trying to push it off the hill.

And all the while, my 5-yr old computer is taking everything I can think to throw at it.

DP

[Fumbili]

What?!? No DP (display port) cables/adapters? The latest and greatest video connector scam. DP to HDMI, DP to DVI, DP to VGA, etc. Ditch WorstBuy and try Monoprice

Osborne Effect? Anyone?

[Chas]

Yeah! We've got this great new tech.

But we have this even better tech coming out just a little later that sucks the doors off, steals it's dog, sleeps with its wife, empties out its bank account, and thoroughly curb-kicks it.

But buy the crappier neat new stuff now!

Historical reference:

• Adam Osborne
• Osborne Computer Corporation
• Osborne 1 Computer

Not enough bandwidth.

[bertok]

Anyone else think that 10Gbps is too little bandwidth for a display interconnect that's not even released yet? Why target the past?

For example, HDMI 1.3 is already at 10.2 Gbps, which is more than Light Peak, and with good reason. For example, Dell has a 27" monitor with Deep Color support, so that's:

    2560h * 1440v * 60Hz * 48 bits per pixel = 10.6 Gbps.

If you want 3D or high framerate gaming with Deep Color even on a smaller 24" screen, you're also out of luck:

    1920h * 1200v * 120Hz * 48 bpp = 13.27 Gbps.

Why target a bandwidth that already can't handle existing displays, when future displays will likely have even higher bandwidths?

Some of the touted features of Light Peak are daisy-chaining and hanging multiple displays off one port. That's just not going to work for any decent modern monitor. Even at the standard 24 bits per pixel, multiple displays won't be possible with two 27" or 30" monitors, or two 24" monitors at 120Hz.

These aren't even high-end professional monitors, Dell will deliver the 27" U2711 for USD 1100 to your door, and 24" monitors that can do 120Hz are common now.

Re:Not enough bandwidth.

[KonoWatakushi]

Or DisplayPort, which provides:

17.28 Gbit/s of video bandwidth, enough for supporting 4 simultaneous 1080p60 displays or 2560 × 1600 × 30 bit @120 Hz

Re:Not enough bandwidth.

[Anonymous Coward]

p60 != 120fps, wouldnt that be p120?

Re:Not enough bandwidth.

[MikeBabcock]

p60 != 120fps, wouldnt that be p120?

Actually, most 120Hz displays are being fed 60Hz signals and either generating intermediate frames algorithmically or displaying the same frames back to back. Look up "MotionFlow" as one such technology.

The result might 'look' like 120p but the signal over the wire was still 60p.

That said, 60p 3D still requires true 120p transmission speeds.

Re:Not enough bandwidth.

[owlstead]

I guess that if there are faster interfaces then there are faster interfaces. But it is like saying hard drives are better than SSD because of size or cost per gigabyte. This interface can use (a probably cheap) long / strong thin & flexible cable. That alone should make such a thing interesting. Think HD-TV on the other side of the room for instance. This cable should easily fit under the carpet or something like that.

That Intel has been so present minded to be able to interleave different protocols. I still have a grey box (queue: painted black) under my desk and I would love a single cable to the monitors and such on top of the desk. If the monitors could be extended to support daisy chaining and eSATA on top of the USB/card reader, I'm all for it. Of course, you could do the same thing with electrical wiring, but for now Intel seems to be the only one to think this out well.

optical interface

[Anonymous Coward]

optical interface would be a perfect match for my optical mouse.

old tech?

[Anonymous Coward]

don't they already signal fairly quickly (too lazy to search wikipedia) in your run-of-the-mill fiber optic cables?

Re:Osborne Effect? Anyone?

[Anonymous Coward]

Light peak is a new connector that's going to start selling potentially this year, and aims to replace USB, HDMI, etc.

The other thing is a long-term research project involving reducing the price and increasing the bandwidth of optical interfaces. They say there'll be applications within a few years, but not what exactly those are; they could be networking applications, not necessarily consumer device interconnects.

Front side bus

[bebilith]

Bugger this external/peripheral connector stuff.

Stick these transceivers on the CPU/Memory/GPU silicon, have them all connected via fibre and get rid of this stupid front side bus bottleneck.

--
B

Re:Front side bus

[godunc]

Stick these transceivers on the CPU/Memory/GPU silicon, have them all connected via fibre and get rid of this stupid front side bus bottleneck.

This is actually the driving application for hybrid Si-InP integration research. (http://engineering.ucsb.edu/bowers/). The biggest limitation of this technology is the inefficiency of the laser. It seems logical that Intel is going after an electronically simpler application like an integrated tranceiver, before attempting to integrate with an already hot running microprocessor.

add some electrons to those photons.

[distantbody]

For this optical interconnect to be as universal as USB, it must provide power + data. Given the size of an optical fibre, Light Peak is the best opportunity to unify data interconnects, from portable device to displays, networking and even internal PC interconnects
I just hope they don't screw it up like usb with a dozen different connectors.
On that note, there will be a new micro connector for USB3, which means another charger and more cables. I'm unsure that Intel will do Light Peak right by the consumer.

Comment removed

[account_deleted]

Comment removed based on user account deletion

I know, right?

[Anonymous Coward]

We should totally make it so there is only one kind of vehicle for everything too. Scrap all the bicycles, sports cars, trucks, tractors, and golf carts, and replace them all with a line of identical vehicles that may or may not be adequate for any particular job, oh wait...

Libraries of Congress

[asukasoryu]

How many Libraries of Congress you ask?

Built on a technology known as silicon photonics, the link has the potential to scale to up to a terabit per second, enough to transfer the contents of a laptop in less than a second or the entire Library of Congress in less than two minutes, according to Justin Rattner, Intel's chief technical officer.

Wait...

[kuei12]

So we should wait a few years to buy anything from Intel because their up and coming technology is obsolete before it hit the stores? Good strategy.

Amanda Seyfried/Julianne Moore love scene? Check!

[Impy the Impiuos Imp]

And here I was runnin' around telling everybody that 5GBs is enough for anybody. I feel like a complete idiot!

LP, now or later?

[Anonymous Coward]

So 10 Gbps LP is vaporware till about a year or two from now, but better vaporware with 50 Gbps will be exiting vapor stage in 3 years.

Why force everyone through two PLUG update cycles, when you can make the 10 forward compatible with the 10?

This is allegedly one of those one plug for all people things, right?

I ask.