The article mentions that lightpeak is simply a transport. Does anyone know the feasibility of implementing say SCSI, or Firewire protocols on top of it?
Because it's simply a physical layer transport, anything should work over it. SCSI, Firewire, you name it. I believe some of the only problems are making sure those protocols work at the respective speeds, but the physical Light Peak layer does allow bitrates that high.
That's directly from intel, hopefully it doesn't change. They state they have no intentions of changing anything like USB, but rather using USB over Light Peak.
SCSI as a protocol (not the interface) is already used EVERYWHERE for mass-storage applications. Firewire, USB, FibreChannel, and SAS all use the SCSI command-set for mass-storage. I think by extension we can assume that Light Peak already speaks SCSI to some extent since it is meant to be, at the very least, a USB replacement.
Did you know all modern CD/DVD drives actually speak SCSI? ATAPI == ATA Packet Interface, which transports SCSI commands over [S|P]ATA. This was a boon for both the CD/DVD drive manufacturers and software creators which had been using SCSI drives for years and years. With ATAPI, only the physical interface changed, not how you talked to the drives.
ATA, as a storage protocol, is really the "odd man out".
It is bit disappointing that Intel keeps postponing these dates. First SB was going to arrive in Q42010 but now it will arrive in early 2011. Light peak was also going to arrive in Q42010 but it will come one year later. Intel has been, for whole this year, aggressively suggesting that it will come this year. Looks like folks at Intel can't keep any promise. A little competition to Intel is needed here badly.
I know I read it somewhere too. But Intel has been saying for sometime that chipset would be available by Q42010 and products will after that. Now it will take one more year, which sucks. But this is not something new, Intel has recently delayed lots of deadlines. I wish if AMD has started offering USB 3.0 chipset then Intel would have hurried up with the Lightpeak.
Probably it has nothing to do with the technology having bugs. At IDF there was a demonstration with a laptop and it worked smoothly. According to you then there are still bugs with USB 3.0 which Intel is currently fixing and could only fix until the next year? I think It has everything to do with the economics. Intel facing no competition in market is doing everything to maximize its profit and not caring about the consumers. Look at the current CPU unlocking plan. The decision to defer Lighpeak and not supporting USB 3.0 until next year has its roots in the economics and not in the technology.
It is interesting that Intel has been so pigheaded about not supporting USB3 in its most recent chipsets; so one has to wonder. Is the real Intel story that they really want LightPeak to take off, they are willing to ship at cost (or lower) to make this happen, and they are desperately worried that USB3 (which is presumably intrinsically cheaper) will take up most of the volume uses for LightPeak, relegating it to expensive and limited usage for a few years at least? Likewise I have to wonder if Apple is in Intel's camp on this, which is why they, similarly, have been really pigheaded about no USB3 on any Macs. I applaud the IDEA of the strategy, but I fear that it will fail, and that LightPeak will be FireWire all over again --- superior, sure, but with equivalent peripherals costing 20% more to no obvious benefit.
This could be solved, to at least some extent, but making sure that LightPeak carries SATA commands, rather than USB commands, to mass storage. That way external drives would get SMART, command queuing, TRIM (if they are SSDs), all the fancy SATA commands that are not passed over USB (because USB mass storage was designed by morons, and apparently has not been updated for USB3). But I have not seen SATA in the list of protocols slated for transport over LightPeak 1.0. Anyone know anything about this? It would be truly pathetic if, in 2011 we have two high speed external bus standards, USB3 and LightPeak, BOTH of which are incapable of actually supporting a modern drive (SMART, TRIM, NCQ) in a decent fashion.
Actually, Light Peak isn't really a USB competitor. Light Peak is protocol-agnostic. It's a physical transport layer. It's the connectors and the cable. Light Peak IS NOT related to the data or the format of the data moving across the wires. So USB, any version of it, would be just one of many protocols that operate on Light Peak. Light Peak is only a competitor of the physical interfaces, not the protocols.
This would allow you to have a Light Peak connector for just about any device, from internal and external hard drives to addon cards (like video and sound cards) to networking devices or computers to connecting to displays. Light Peak is 10Gbps of cable. NOT protocol.
This retarded comment system thinks my comment is spam, so I have to split it into multiple posts.
(a) LightPeak does the same thing (connect hard drives to computers in fast way) as USB3. That's the only market that matters, and as far as that goes they are competitors. The fact that you can run USB3 protocols over LP hardware is nice for OS programmers, and of no relevance to anyone else. What matters to ye average person is - what is the extra cost of an LP equipped PC vs a USB3 equipped PC and - what is the extra cost of an LP equipped HD vs a USB3 equipped HD.
(b) The fact that LP can run SCSI or FiberChannel or whatever is of zero interest to the volume market. The fact that it can run ethernet is of marginal interest given that everyone already has a stash of ethernet cables, hubs, etc, all of which work adequately.
This is why I made such a big deal about ATA over LP --- that is the one differentiating factor from USB3 that actually makes an important difference, in the one market that matters. If LP can position itself as "yes, we admit our HDs cost $20 more than USB3 HDs, but they do give you SMART, NCQ and TRIM, and that's of some value", it has a chance. If it can't even offer that, it's FW vs USB2 all over again --- another situation where FW was too damn stupid to appreciate that it should embrace native ATA commands over the physical bus rather than this stupid SCSI emulation crap. (obRant This fetishization of command sets from thirty years ago is a really stupid and embarrassing aspect of contemporary computer HW. It's stupid in the case of USB/FW mass storage, and it's stupid in the case of Bluetooth emulating modems. I'm not crazy about it in the case of ATA either, but if you're going to hitch your wagon to a has-been, at least choose a has-been of some relevance to the modern world.)
Actually the protocol DOES matter to a consumer. Different protocols, regardless of physical transport, require licensing to use. This licensing is what determines, in part, the cost of devices and peripherals. The cost of computers with LP would probably be no different than any other bleeding edge technology at first, but come down quickly. However, the cost of devices and peripherals would be based on the licensed protocols.
The fact that LP runs ANY protocol over its cable means a device manufacturer can put an LP connector on their product and use a cheap protocol, like USB. In fact, licensing is probably the reason for all the convoluted protocol issues with USB. If a manufacturer can use LP for the connector and whatever protocol is best for the task, they can custom tailor their protocol choice to whatever best fits their devices features. A dumb mass storage drive might go with USB simply because SMART, NCQ, TRIM, et. al are overkill. But if you slap a LP connector on an external SSD and use SATA as the protocol, you could have a very efficient storage device without having to open your computer or take a drive out of a hot-swap bay.
LP isn't entirely a USB competitor. It's a USB cable competitor. The USB-IF will probably still get a licensing fee and of course Intel will get a piece. Of course, manufacturers could stick with USB instead of investing in LP at all and ride the bargain train.
How is the power being transmitted here? I realize they must have used some sort of copper cable to provide the power but there is no detail on this. Would be good to know.
I'm going to follow up with Intel and ask about power transport. I believe it is possible and there's a copper wire alongside the waveguide, but I'm going to check.
A whole article on Light Peak and nothing related to Apple? Far as I had heard, Apple helped a ton with creation of this technology and is first in line to adopt into their whole product range. Check out the link below with diagrams etc -
wouldn't be surprised if Apple drove the whole lightpeak initiative as the technology choice here makes no sense.
why would intel use VCSEL laser? those usually have good linewidth characteristics but a lot more expensive than DFBs and other cheaper types. For 10Gbps, you can easily directly modulate a DFB and get good reach. I suspect with VCSEL laser, there's a need for an external modulator which makes it a lot more expensive. Also on the topic of WDM, right now ppl are doing >40Gbps and 100Gbps with single wavelength, going wdm with 10gig is just a bad choice. The use of multimode fiber is not the best either. Corning already has bendable (very tight radius) single mode fiber, going multimode just create extra problems with modal dispersion and definitely requires a better receiver -> more expensive.
Hmm. So let me get this right. You hate the choice of a VCSEL laser because DFB would be cheaper (but not as good tech). And you hate the choice of multimode even though IT is cheaper than single mode, because it is not as good tech.
Why not just come out straight and say that you hate this because it was spearheaded by Apple, nothing to do with the details of the tech, you just hate anything associated with Apple? Don't waste our time with your pseudo-scientific teardown of the tech when those details have fsckall to do with your actual feelings.
I actually work in optical communication industry and I build optical modems. Yes DFB is cheaper than VCSEL mainly b/c it's been around a lot longer and the manufacturing process is a lot more refined. Multimode is actually quite a bit more expensive than single mode, simply b/c it needs more glass. Ask anyone from corning if you don't believe me. I never said multimode is cheaper, I did say a more expensive receiver is required if you use multimode fiber. So essentially every component choice made in Light Peak is expensive. Is that enough to convince you or your ego?
As to RaynorWolfcastle's comment, connector alignment with SMF is really no issue with the standardization of SC or FC connector, and the bendable single mode fiber is still cheaper than the equivalent multimode. Only real problem is keeping the SMF core clean to get enough signal sensitivity.
Intel uses directly modulated VCSELs, though I can't remember where I read this (Light Reading, maybe?). I believe that VCSELs are also more die-space efficient than DFBs since they have (by definition) a vertical cavity vs the DFB's horizontal cavity.
Secondly, it is very obvious that they would want multimode fiber for easier alignment and cheaper connectors. Dispersion is not really a concern over such short distances (tens of meters) anyways. Corning's photonic crystal fiber is also expensive.
You're really looking at this from a performance point of view instead of looking at it from a cost-reduction point of view. They're not looking to generate clean PSK signals over kms of fiber here, they want cheap connectors with cheap cables that can be used without any fuss by a sexagenarian.
As an aside, I really think the connector issue is a much bigger one than the article lets on. Making connectors that align well for pennies will be challenging.
No offense, but was the author half-asleep while writing this? I see many grammar errors in this piece...Frankly, the writing in this article is lower than what I usually see and expect from Anandtech.
"All together, the vision is enabling creation enabling the Compal notebook to control mixing on the Avid HD I/O box."
"All total, Light Peak carried uncompressed audio mixed..."
In theory it can work. But what are the chances of this causing an aftermarket vendor to make external video cards? This would be great both for laptops (so you can have battery life and speed when you need) and make it less scary for people to upgrade their video cards in their desktops.
ATA is in. USB is in. Any protocol is in. It's up to the device to determine the protocol. You could have external drives using SCSI, USB, or something of their own proprietary design. That's the biggest point of LP: It doesn't care how you organize your bits, it just wants to get them there fast.
could this finally boot expensive fiber channel and provide an inexpensive method for enterprise SAN connections at a far lower cost than FC or 10gbe iscsi?
The goals that this is set to achieve are: 1. 300ft point-to-point cables 2. One cable type for multiple protocols
So...who cares? Repeaters or wireless can already give us cables as long as we want. And what's the point of 1 cable type for all devices? This won't reduce the number of cables involved, it just means 1 type. I don't care what type connects my PC to my monitor, do you? I also think Intel's claim that this lowers cost is bunk, since this means you'll need Light Peak bridges all over the place to convert to our established protocols. That also means that every component would need to adopt it, which won't happen anytime soon.
And why get light involved? If you want a single cable with tons of bandwidth, just use multi-lane PCIe. That'd probably be a lot cheaper, too. USB3 is like 1 lane PCIe2, and they probably could have made it more lanes, but that would increase the cost for unnecessary bandwidth.
I'm with ya, JoKeRr. Apple's pushing it cuz it's "oooo, light" and their next big idiotic marketing thing.
Best Buy charges $230 for a 6.6ft HDMI cable. Who wants to pay $10,000 for a 300ft optical?
Aside for having one type of cable to most devices, fiber optics is just the way to go for more bandwidth rather than making more copper lines/wires.
One thing I would agree on is this won't make it mainstream in the near future. USB 3.0 is hugely fast for almost all devices and provides electricity to a device, yet I could see Light Peak being used for display connections, HD cameras, external drives in one or two years.
One cable type means you'll have a bunch of ONE type of peripheral connector on your computer that can connect to ANY Light Peak peripheral. It doesn't care what protocol it uses, just the connector needs to be right. I really like the idea of having a dozen or so Light Peak connectors on my computer and just plugging in devices without a second thought. I don't like have a dozen connectors, only four of each variety and needing more of one type.
Fiberoptic is the only economical way to achieve the kind of bandwidth needs we're moving toward and keep cable lengths long. Remember, Intel already demonstrated a 100Gbps version. That kind of cable could handle just about anything you can throw at it, making it perfect for everything from an audo connector, display connector, or for mass storage devices. I'd love to be able to pick up an external drive with a couple TB of storage in a RAID and be able to feed it data as fast as it could write it. And then use the same connector to hook up my HD camera to unload videos as fast as it can read them.
"I'd love to be able to pick up an external drive with a couple TB of storage in a RAID and be able to feed it data as fast as it could write it. And then use the same connector to hook up my HD camera to unload videos as fast as it can read them."
Well good, then you'd love to use the USB standard!
You brought up the issue of speed, as if LP enables higher transfer speeds. It doesn't, as it uses existing protocols. And you said audio cable--that will remain analog for most people.
To help you understand, consider this. DVI has many parallel data pins. These would be fed into a DVI-LP bridge, light would be transmitted through the cable, and there'd be another LP-DVI bridge in the monitor, where it's then spread out to the parallel pins that the monitor system is ready to handle. LP adds no speed.
That is not enough for DisplayPort 1.2. Not to mention the higher resolution monitor that are coming out when Lightpeak are available, and the need for 120hz 3D Display.
While i would love a new and improved connector. ( I hate the way USB allows you to plug in upside down ). Sticking to USB connector would make more sense given the huge amount of devices using it such as USB memory. The Cooper left inside could be used to power transmission.
Huh? There's a barrel maker inside each of my USB connectors? And now you are suggesting they can be put to work with LightPeak as well? Truly amazing!
Seriously though, the confusion for many that would result from having USB connectors that aren't actually USB connectors would be too much. I doubt even color coding would suffice.
"Interestingly, Intel is using an 850 nm wavelength laser for the system"
This doesn't seem that unusual when you consider that 850nm multi-mode fiber is commonly used for 10gig Ethernet (e.g. XFP modules), so it's probably readily available and cheap. The fact that it's not entirely eye-safe is kind of interesting though for a consumer-grade product, though I'm sure it's running at very low power. I guess the fact that the entire optical path is sealed inside the cable makes it safe enough, unless you go staring into the end of a broken cable.
Because this is an Intel product, they may decide to cut AMD out of the picture. If AMD can't get lightpeak into its chipsets, and its as good as they say it is, that would be a heavy blow. Hopefully they can't extend that too far and will allow 3rd party chips like USB3 is now on mobos... Otherwise all I can do is yell "monopoly"...
"The views expressed here are mine and do not reflect the official opinion of my employer or the organization through which the Internet was accessed."
All the early LightPeak connectors are based on USB, which gave me great hope that LightPeak would be backwards compatible with USB. After all, Intel's intention is to replace all connectors with LightPeak... using a new connector will render every USB device on the market incompatible.
Yes, I realize that there is a transition period here, where we'll get a few LightPeak connectors to start out with, but breaking backwards compatibility is just plain dumb. For one thing, some peripherals just don't NEED optical connections. Does my keyboard really need a 10Gbps optical connection, driving up costs? No! Maintaining backwards compatibility with USB would have not only allowed all existing peripherals to continue working, it would have allowed lower cost devices to avoid using optical when they don't need it.
Look how long it's taken us to transition from PCI to PCIe. PCI-Express was introduced in 2004. SIX years ago. And even today, most motherboards still ship with regular PCI slots because there are still devices that use PCI on the market. PCI NICs are still cheaper than PCIe NICs.
Now, PCI was never nearly as widely used as USB. Many computers didn't even have any PCI cards, since lower-end motherboards tended to build video/audio/etc into the chipset. On the other hand, everybody out there has multiple USB devices plugged into their computer. If we still haven't fully transitioned off PCI six years later, can you imagine how long killing off USB is going to take?
This was such a great opportunity for Intel to create the successor to USB. Instead, they've decided to create the usurper. And that's very unfortunate.
The first generation LP will have competition from USB 3. For one thing, USB 3 is instantly backward compatible with the huge installed base of existing USB 2 devices. LP would need some sort of adapter and AFAIK doesn't provide power. So USB 3 will be much cheaper and at 10x the USB 2 bandwidth will seem "fast enough" for a few years.
The big question, right now, is not Intel working LP into a chipset, or even into a SoC, but rather what about the device interface chip? Can drive makers selling $80 USB drives add a LP interface, or does that make the $80 drive a $120 drive? The USB 3 will already transfer data far faster than the drive can keep up with. The LP device interface chip has to be cost cometitive with the USB 3 chip.
However, I see the big move coming. When their integrated photonics stuff gets rolling, the cost of LP will dramatically fall. There is no question that the LP bandwidth has much room for growth, while any copper cabling system has near zero room for growth. Optical transmission is definitely the future, but the cost has to come down and something has to be done about supplying power for external devices.
M$ not have possibilities to make a world size standart of comunication , if they come out of full support of the other biggest trademark in this industry!
Fails to deliver power over the connection. Thus ultimately it is about 1/10th as useful to the general population who will vote for simplicity over speed. USB 3 gives you both speed and power and will be multiples more useful.
If they can keep this stuff relatively cheap, the 50m Distance along with the ability to Daisy chain for longer distances means this would make a pretty decent Home Networking medium.
How would you like to have 10GE strung around your entire house?
Or for that matter, take your HDD out of your desktop and have it in the basement.
Or hell, LPK enabled Monitor and leave the entire PC in the basement.
Can't wait to see what level they take this to and I pray they keep it fairly cheap.
"obviously the connector will be subject to the same kind of scrutiny that USB was for durability and consumer friendliness"
I hope not! While definitely nicer than old-style long-pin connectors, trying to plug a USB connector in without squinting at it and the plug fails half the time due to being upside down. Hope they can take a clue from Firewire-400 here.
"but the display itself was only 1080p"? WTF? You guys are really starting to piss me off. Resolutions over 1080p are useless for most people, many people still don't even have 1080p on all their screens. Also, I see almost no point in going to 1440p or 1600p, it's not going to be enough to make a noticeable difference; diminishing returns and all. Let the world catch up to 1080p, then, sometime around 2020, we can start getting excited about 16,160X9,090 ie 9090p or whatever becomes the next big step. Quite frankly 4320p isn't enough to convince me to spend any extra money, I'd still wait for a 37" to drop down under 500 bucks. 3D and stupid widgets I'll never use and all. /rant.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
59 Comments
Back to Article
gft74 - Thursday, September 16, 2010 - link
The article mentions that lightpeak is simply a transport. Does anyone know the feasibility of implementing say SCSI, or Firewire protocols on top of it?-gft
vol7ron - Thursday, September 16, 2010 - link
that shouldn't be hard, but really why would you want to?the real aim should be that external devices begin to have a light port.
Brian Klug - Thursday, September 16, 2010 - link
Because it's simply a physical layer transport, anything should work over it. SCSI, Firewire, you name it. I believe some of the only problems are making sure those protocols work at the respective speeds, but the physical Light Peak layer does allow bitrates that high.That's directly from intel, hopefully it doesn't change. They state they have no intentions of changing anything like USB, but rather using USB over Light Peak.
-Brian
cdillon - Thursday, September 16, 2010 - link
SCSI as a protocol (not the interface) is already used EVERYWHERE for mass-storage applications. Firewire, USB, FibreChannel, and SAS all use the SCSI command-set for mass-storage. I think by extension we can assume that Light Peak already speaks SCSI to some extent since it is meant to be, at the very least, a USB replacement.Did you know all modern CD/DVD drives actually speak SCSI? ATAPI == ATA Packet Interface, which transports SCSI commands over [S|P]ATA. This was a boon for both the CD/DVD drive manufacturers and software creators which had been using SCSI drives for years and years. With ATAPI, only the physical interface changed, not how you talked to the drives.
ATA, as a storage protocol, is really the "odd man out".
solipsism - Thursday, September 16, 2010 - link
Is the adapter receiving power from a cooper power wire from LightPeak source?holywarrior007 - Thursday, September 16, 2010 - link
It is bit disappointing that Intel keeps postponing these dates. First SB was going to arrive in Q42010 but now it will arrive in early 2011. Light peak was also going to arrive in Q42010 but it will come one year later. Intel has been, for whole this year, aggressively suggesting that it will come this year. Looks like folks at Intel can't keep any promise. A little competition to Intel is needed here badly.melgross - Thursday, September 16, 2010 - link
Intel has just said that we won't see any products until 2012 because the chipset won't be ready until late 2011.holywarrior007 - Thursday, September 16, 2010 - link
I know I read it somewhere too. But Intel has been saying for sometime that chipset would be available by Q42010 and products will after that. Now it will take one more year, which sucks. But this is not something new, Intel has recently delayed lots of deadlines. I wish if AMD has started offering USB 3.0 chipset then Intel would have hurried up with the Lightpeak.vol7ron - Thursday, September 16, 2010 - link
I rather see delays then products with complications.probedb - Friday, September 17, 2010 - link
Well they can either delay it and get it right or rush it out to make you happy and it could be full of bugs.I know which I'd prefer.
holywarrior007 - Tuesday, September 21, 2010 - link
Probably it has nothing to do with the technology having bugs. At IDF there was a demonstration with a laptop and it worked smoothly. According to you then there are still bugs with USB 3.0 which Intel is currently fixing and could only fix until the next year? I think It has everything to do with the economics. Intel facing no competition in market is doing everything to maximize its profit and not caring about the consumers. Look at the current CPU unlocking plan. The decision to defer Lighpeak and not supporting USB 3.0 until next year has its roots in the economics and not in the technology.name99 - Thursday, September 16, 2010 - link
It is interesting that Intel has been so pigheaded about not supporting USB3 in its most recent chipsets; so one has to wonder. Is the real Intel story that they really want LightPeak to take off, they are willing to ship at cost (or lower) to make this happen, and they are desperately worried that USB3 (which is presumably intrinsically cheaper) will take up most of the volume uses for LightPeak, relegating it to expensive and limited usage for a few years at least?Likewise I have to wonder if Apple is in Intel's camp on this, which is why they, similarly, have been really pigheaded about no USB3 on any Macs.
I applaud the IDEA of the strategy, but I fear that it will fail, and that LightPeak will be FireWire all over again --- superior, sure, but with equivalent peripherals costing 20% more to no obvious benefit.
This could be solved, to at least some extent, but making sure that LightPeak carries SATA commands, rather than USB commands, to mass storage. That way external drives would get SMART, command queuing, TRIM (if they are SSDs), all the fancy SATA commands that are not passed over USB (because USB mass storage was designed by morons, and apparently has not been updated for USB3).
But I have not seen SATA in the list of protocols slated for transport over LightPeak 1.0. Anyone know anything about this? It would be truly pathetic if, in 2011 we have two high speed external bus standards, USB3 and LightPeak, BOTH of which are incapable of actually supporting a modern drive (SMART, TRIM, NCQ) in a decent fashion.
jordanclock - Friday, September 17, 2010 - link
Actually, Light Peak isn't really a USB competitor. Light Peak is protocol-agnostic. It's a physical transport layer. It's the connectors and the cable. Light Peak IS NOT related to the data or the format of the data moving across the wires. So USB, any version of it, would be just one of many protocols that operate on Light Peak. Light Peak is only a competitor of the physical interfaces, not the protocols.This would allow you to have a Light Peak connector for just about any device, from internal and external hard drives to addon cards (like video and sound cards) to networking devices or computers to connecting to displays. Light Peak is 10Gbps of cable. NOT protocol.
name99 - Friday, September 17, 2010 - link
This retarded comment system thinks my comment is spam, so I have to split it into multiple posts.(a)
LightPeak does the same thing (connect hard drives to computers in fast way) as USB3. That's the only market that matters, and as far as that goes they are competitors. The fact that you can run USB3 protocols over LP hardware is nice for OS programmers, and of no relevance to anyone else. What matters to ye average person is
- what is the extra cost of an LP equipped PC vs a USB3 equipped PC and
- what is the extra cost of an LP equipped HD vs a USB3 equipped HD.
name99 - Friday, September 17, 2010 - link
(b) The fact that LP can run SCSI or FiberChannel or whatever is of zero interest to the volume market. The fact that it can run ethernet is of marginal interest given that everyone already has a stash of ethernet cables, hubs, etc, all of which work adequately.This is why I made such a big deal about ATA over LP --- that is the one differentiating factor from USB3 that actually makes an important difference, in the one market that matters. If LP can position itself as "yes, we admit our HDs cost $20 more than USB3 HDs, but they do give you SMART, NCQ and TRIM, and that's of some value", it has a chance. If it can't even offer that, it's FW vs USB2 all over again --- another situation where FW was too damn stupid to appreciate that it should embrace native ATA commands over the physical bus rather than this stupid SCSI emulation crap. (obRant This fetishization of command sets from thirty years ago is a really stupid and embarrassing aspect of contemporary computer HW. It's stupid in the case of USB/FW mass storage, and it's stupid in the case of Bluetooth emulating modems. I'm not crazy about it in the case of ATA either, but if you're going to hitch your wagon to a has-been, at least choose a has-been of some relevance to the modern world.)
jordanclock - Friday, September 17, 2010 - link
Actually the protocol DOES matter to a consumer. Different protocols, regardless of physical transport, require licensing to use. This licensing is what determines, in part, the cost of devices and peripherals. The cost of computers with LP would probably be no different than any other bleeding edge technology at first, but come down quickly. However, the cost of devices and peripherals would be based on the licensed protocols.The fact that LP runs ANY protocol over its cable means a device manufacturer can put an LP connector on their product and use a cheap protocol, like USB. In fact, licensing is probably the reason for all the convoluted protocol issues with USB. If a manufacturer can use LP for the connector and whatever protocol is best for the task, they can custom tailor their protocol choice to whatever best fits their devices features. A dumb mass storage drive might go with USB simply because SMART, NCQ, TRIM, et. al are overkill. But if you slap a LP connector on an external SSD and use SATA as the protocol, you could have a very efficient storage device without having to open your computer or take a drive out of a hot-swap bay.
LP isn't entirely a USB competitor. It's a USB cable competitor. The USB-IF will probably still get a licensing fee and of course Intel will get a piece. Of course, manufacturers could stick with USB instead of investing in LP at all and ride the bargain train.
XZerg - Thursday, September 16, 2010 - link
How is the power being transmitted here? I realize they must have used some sort of copper cable to provide the power but there is no detail on this. Would be good to know.Brian Klug - Thursday, September 16, 2010 - link
I'm going to follow up with Intel and ask about power transport. I believe it is possible and there's a copper wire alongside the waveguide, but I'm going to check.-Brian
tdtd - Thursday, September 16, 2010 - link
A whole article on Light Peak and nothing related to Apple? Far as I had heard, Apple helped a ton with creation of this technology and is first in line to adopt into their whole product range. Check out the link below with diagrams etc -http://www.engadget.com/2009/09/26/exclusive-apple...
tlmaclennan - Thursday, September 16, 2010 - link
There may not have been any mention of Apple in the article, but it is clear that there is a picture of a MacBook Pro with the Light Peak technology.JoKeRr - Thursday, September 16, 2010 - link
wouldn't be surprised if Apple drove the whole lightpeak initiative as the technology choice here makes no sense.why would intel use VCSEL laser? those usually have good linewidth characteristics but a lot more expensive than DFBs and other cheaper types. For 10Gbps, you can easily directly modulate a DFB and get good reach. I suspect with VCSEL laser, there's a need for an external modulator which makes it a lot more expensive. Also on the topic of WDM, right now ppl are doing >40Gbps and 100Gbps with single wavelength, going wdm with 10gig is just a bad choice. The use of multimode fiber is not the best either. Corning already has bendable (very tight radius) single mode fiber, going multimode just create extra problems with modal dispersion and definitely requires a better receiver -> more expensive.
ya smells like a bad apple to me.
name99 - Thursday, September 16, 2010 - link
Hmm. So let me get this right.You hate the choice of a VCSEL laser because DFB would be cheaper (but not as good tech).
And you hate the choice of multimode even though IT is cheaper than single mode, because it is not as good tech.
Why not just come out straight and say that you hate this because it was spearheaded by Apple, nothing to do with the details of the tech, you just hate anything associated with Apple? Don't waste our time with your pseudo-scientific teardown of the tech when those details have fsckall to do with your actual feelings.
JoKeRr - Friday, September 17, 2010 - link
I actually work in optical communication industry and I build optical modems. Yes DFB is cheaper than VCSEL mainly b/c it's been around a lot longer and the manufacturing process is a lot more refined. Multimode is actually quite a bit more expensive than single mode, simply b/c it needs more glass. Ask anyone from corning if you don't believe me. I never said multimode is cheaper, I did say a more expensive receiver is required if you use multimode fiber. So essentially every component choice made in Light Peak is expensive. Is that enough to convince you or your ego?As to RaynorWolfcastle's comment, connector alignment with SMF is really no issue with the standardization of SC or FC connector, and the bendable single mode fiber is still cheaper than the equivalent multimode. Only real problem is keeping the SMF core clean to get enough signal sensitivity.
RaynorWolfcastle - Thursday, September 16, 2010 - link
Intel uses directly modulated VCSELs, though I can't remember where I read this (Light Reading, maybe?). I believe that VCSELs are also more die-space efficient than DFBs since they have (by definition) a vertical cavity vs the DFB's horizontal cavity.Secondly, it is very obvious that they would want multimode fiber for easier alignment and cheaper connectors. Dispersion is not really a concern over such short distances (tens of meters) anyways. Corning's photonic crystal fiber is also expensive.
You're really looking at this from a performance point of view instead of looking at it from a cost-reduction point of view. They're not looking to generate clean PSK signals over kms of fiber here, they want cheap connectors with cheap cables that can be used without any fuss by a sexagenarian.
As an aside, I really think the connector issue is a much bigger one than the article lets on. Making connectors that align well for pennies will be challenging.
DigitalFreak - Thursday, September 16, 2010 - link
"A whole article on Light Peak and nothing related to Apple?"Waaaaaaa
Stahn Aileron - Thursday, September 16, 2010 - link
No offense, but was the author half-asleep while writing this? I see many grammar errors in this piece...Frankly, the writing in this article is lower than what I usually see and expect from Anandtech."All together, the vision is enabling creation enabling the Compal notebook to control mixing on the Avid HD I/O box."
"All total, Light Peak carried uncompressed audio mixed..."
Brian Klug - Thursday, September 16, 2010 - link
I actually was half asleep, around 1:00 AM ;)Fixing those now!
-Brian
numberoneoppa - Thursday, September 16, 2010 - link
Haha, Brian. <3SteelCity1981 - Thursday, September 16, 2010 - link
So this is why Intel has kept putting native USB 3.0 support on the back burner.Roland00 - Thursday, September 16, 2010 - link
In theory it can work. But what are the chances of this causing an aftermarket vendor to make external video cards? This would be great both for laptops (so you can have battery life and speed when you need) and make it less scary for people to upgrade their video cards in their desktops.digitaldreamer - Thursday, September 16, 2010 - link
What's the story on ATA? Or, is that assumed to be included as part of the other protocols?MAJ
digitaldreamer - Thursday, September 16, 2010 - link
And, what protocol was used for the external hard drive? USB?MAJ
<wishing for a EDIT option>
jordanclock - Friday, September 17, 2010 - link
ATA is in. USB is in. Any protocol is in. It's up to the device to determine the protocol. You could have external drives using SCSI, USB, or something of their own proprietary design. That's the biggest point of LP: It doesn't care how you organize your bits, it just wants to get them there fast.Mike1111 - Thursday, September 16, 2010 - link
All the LightPeak prototypes seem to have two optical fibers. Wouldn't that mean 20 Gbit/s in each direction?SimKill - Thursday, September 16, 2010 - link
I think its bidirectional, i.e. one for Transmit and one for Receive.So each line would still have only 10Gbps
J_Tarasovic - Thursday, September 16, 2010 - link
Nice work. I am pretty excited about the technology. Still waiting on some SSD news though.DigitalFreak - Thursday, September 16, 2010 - link
I was beginning to think someone from DailyTech snuck an article in. :0)pesos - Thursday, September 16, 2010 - link
could this finally boot expensive fiber channel and provide an inexpensive method for enterprise SAN connections at a far lower cost than FC or 10gbe iscsi?AnnonymousCoward - Thursday, September 16, 2010 - link
The goals that this is set to achieve are:1. 300ft point-to-point cables
2. One cable type for multiple protocols
So...who cares? Repeaters or wireless can already give us cables as long as we want. And what's the point of 1 cable type for all devices? This won't reduce the number of cables involved, it just means 1 type. I don't care what type connects my PC to my monitor, do you? I also think Intel's claim that this lowers cost is bunk, since this means you'll need Light Peak bridges all over the place to convert to our established protocols. That also means that every component would need to adopt it, which won't happen anytime soon.
And why get light involved? If you want a single cable with tons of bandwidth, just use multi-lane PCIe. That'd probably be a lot cheaper, too. USB3 is like 1 lane PCIe2, and they probably could have made it more lanes, but that would increase the cost for unnecessary bandwidth.
I'm with ya, JoKeRr. Apple's pushing it cuz it's "oooo, light" and their next big idiotic marketing thing.
Best Buy charges $230 for a 6.6ft HDMI cable. Who wants to pay $10,000 for a 300ft optical?
zodiacfml - Thursday, September 16, 2010 - link
Aside for having one type of cable to most devices, fiber optics is just the way to go for more bandwidth rather than making more copper lines/wires.One thing I would agree on is this won't make it mainstream in the near future. USB 3.0 is hugely fast for almost all devices and provides electricity to a device, yet I could see Light Peak being used for display connections, HD cameras, external drives in one or two years.
AnnonymousCoward - Friday, September 17, 2010 - link
"fiber optics is just the way to go"No, COST is just the way to go. Note that whoever created DisplayPort chose to go the route of up to 4 PCIe lanes rather than fiber optics.
jordanclock - Friday, September 17, 2010 - link
One cable type means you'll have a bunch of ONE type of peripheral connector on your computer that can connect to ANY Light Peak peripheral. It doesn't care what protocol it uses, just the connector needs to be right. I really like the idea of having a dozen or so Light Peak connectors on my computer and just plugging in devices without a second thought. I don't like have a dozen connectors, only four of each variety and needing more of one type.Fiberoptic is the only economical way to achieve the kind of bandwidth needs we're moving toward and keep cable lengths long. Remember, Intel already demonstrated a 100Gbps version. That kind of cable could handle just about anything you can throw at it, making it perfect for everything from an audo connector, display connector, or for mass storage devices. I'd love to be able to pick up an external drive with a couple TB of storage in a RAID and be able to feed it data as fast as it could write it. And then use the same connector to hook up my HD camera to unload videos as fast as it can read them.
AnnonymousCoward - Friday, September 17, 2010 - link
"I'd love to be able to pick up an external drive with a couple TB of storage in a RAID and be able to feed it data as fast as it could write it. And then use the same connector to hook up my HD camera to unload videos as fast as it can read them."Well good, then you'd love to use the USB standard!
You brought up the issue of speed, as if LP enables higher transfer speeds. It doesn't, as it uses existing protocols. And you said audio cable--that will remain analog for most people.
AnnonymousCoward - Friday, September 17, 2010 - link
To help you understand, consider this. DVI has many parallel data pins. These would be fed into a DVI-LP bridge, light would be transmitted through the cable, and there'd be another LP-DVI bridge in the monitor, where it's then spread out to the parallel pins that the monitor system is ready to handle. LP adds no speed.iwodo - Thursday, September 16, 2010 - link
That is not enough for DisplayPort 1.2. Not to mention the higher resolution monitor that are coming out when Lightpeak are available, and the need for 120hz 3D Display.While i would love a new and improved connector. ( I hate the way USB allows you to plug in upside down ). Sticking to USB connector would make more sense given the huge amount of devices using it such as USB memory. The Cooper left inside could be used to power transmission.
wisecracker - Saturday, September 18, 2010 - link
Huh? There's a barrel maker inside each of my USB connectors? And now you are suggesting they can be put to work with LightPeak as well? Truly amazing!Seriously though, the confusion for many that would result from having USB connectors that aren't actually USB connectors would be too much. I doubt even color coding would suffice.
ravin511 - Friday, September 17, 2010 - link
The power is coming from the HDMI cable. The HDMI cable has 5V power inside it.Metaluna - Friday, September 17, 2010 - link
"Interestingly, Intel is using an 850 nm wavelength laser for the system"This doesn't seem that unusual when you consider that 850nm multi-mode fiber is commonly used for 10gig Ethernet (e.g. XFP modules), so it's probably readily available and cheap. The fact that it's not entirely eye-safe is kind of interesting though for a consumer-grade product, though I'm sure it's running at very low power. I guess the fact that the entire optical path is sealed inside the cable makes it safe enough, unless you go staring into the end of a broken cable.
gnesterenko - Friday, September 17, 2010 - link
Because this is an Intel product, they may decide to cut AMD out of the picture. If AMD can't get lightpeak into its chipsets, and its as good as they say it is, that would be a heavy blow. Hopefully they can't extend that too far and will allow 3rd party chips like USB3 is now on mobos... Otherwise all I can do is yell "monopoly"..."The views expressed here are mine and do not reflect the official opinion of my employer or the organization through which the Internet was accessed."
Guspaz - Friday, September 17, 2010 - link
All the early LightPeak connectors are based on USB, which gave me great hope that LightPeak would be backwards compatible with USB. After all, Intel's intention is to replace all connectors with LightPeak... using a new connector will render every USB device on the market incompatible.Yes, I realize that there is a transition period here, where we'll get a few LightPeak connectors to start out with, but breaking backwards compatibility is just plain dumb. For one thing, some peripherals just don't NEED optical connections. Does my keyboard really need a 10Gbps optical connection, driving up costs? No! Maintaining backwards compatibility with USB would have not only allowed all existing peripherals to continue working, it would have allowed lower cost devices to avoid using optical when they don't need it.
Look how long it's taken us to transition from PCI to PCIe. PCI-Express was introduced in 2004. SIX years ago. And even today, most motherboards still ship with regular PCI slots because there are still devices that use PCI on the market. PCI NICs are still cheaper than PCIe NICs.
Now, PCI was never nearly as widely used as USB. Many computers didn't even have any PCI cards, since lower-end motherboards tended to build video/audio/etc into the chipset. On the other hand, everybody out there has multiple USB devices plugged into their computer. If we still haven't fully transitioned off PCI six years later, can you imagine how long killing off USB is going to take?
This was such a great opportunity for Intel to create the successor to USB. Instead, they've decided to create the usurper. And that's very unfortunate.
Jaybus - Friday, September 17, 2010 - link
The first generation LP will have competition from USB 3. For one thing, USB 3 is instantly backward compatible with the huge installed base of existing USB 2 devices. LP would need some sort of adapter and AFAIK doesn't provide power. So USB 3 will be much cheaper and at 10x the USB 2 bandwidth will seem "fast enough" for a few years.The big question, right now, is not Intel working LP into a chipset, or even into a SoC, but rather what about the device interface chip? Can drive makers selling $80 USB drives add a LP interface, or does that make the $80 drive a $120 drive? The USB 3 will already transfer data far faster than the drive can keep up with. The LP device interface chip has to be cost cometitive with the USB 3 chip.
However, I see the big move coming. When their integrated photonics stuff gets rolling, the cost of LP will dramatically fall. There is no question that the LP bandwidth has much room for growth, while any copper cabling system has near zero room for growth. Optical transmission is definitely the future, but the cost has to come down and something has to be done about supplying power for external devices.
wwwcd - Friday, September 17, 2010 - link
M$ not have possibilities to make a world size standart of comunication , if they come out of full support of the other biggest trademark in this industry!FXi - Friday, September 17, 2010 - link
Fails to deliver power over the connection. Thus ultimately it is about 1/10th as useful to the general population who will vote for simplicity over speed. USB 3 gives you both speed and power and will be multiples more useful.Pwnx0r - Friday, September 17, 2010 - link
There's going to be a copper wire in the cable for power.boden - Saturday, September 18, 2010 - link
One cable to rule them all,One cable to find them,
One cable to bring them all
and in the light, bind them.
Casper42 - Saturday, September 18, 2010 - link
If they can keep this stuff relatively cheap, the 50m Distance along with the ability to Daisy chain for longer distances means this would make a pretty decent Home Networking medium.How would you like to have 10GE strung around your entire house?
Or for that matter, take your HDD out of your desktop and have it in the basement.
Or hell, LPK enabled Monitor and leave the entire PC in the basement.
Can't wait to see what level they take this to and I pray they keep it fairly cheap.
ABR - Monday, September 20, 2010 - link
"obviously the connector will be subject to the same kind of scrutiny that USB was for durability and consumer friendliness"I hope not! While definitely nicer than old-style long-pin connectors, trying to plug a USB connector in without squinting at it and the plug fails half the time due to being upside down. Hope they can take a clue from Firewire-400 here.
bobbozzo - Monday, September 20, 2010 - link
One thing I haven't seen mentioned is if LP controllers will use DMA... this would be a big advantage over USB.Bob
Hrel - Wednesday, September 29, 2010 - link
"but the display itself was only 1080p"? WTF? You guys are really starting to piss me off. Resolutions over 1080p are useless for most people, many people still don't even have 1080p on all their screens. Also, I see almost no point in going to 1440p or 1600p, it's not going to be enough to make a noticeable difference; diminishing returns and all. Let the world catch up to 1080p, then, sometime around 2020, we can start getting excited about 16,160X9,090 ie 9090p or whatever becomes the next big step. Quite frankly 4320p isn't enough to convince me to spend any extra money, I'd still wait for a 37" to drop down under 500 bucks. 3D and stupid widgets I'll never use and all. /rant.