Slashdot Mirror
Production of Photon Processors Expected in 2006
ThinSkin writes "Photon processors that transmit data via light, not electrons, are slated to enter production in mid-2006, ExtremeTech reports. Headed by a UCLA professor and a Nobel Prize winner, startup Luxtera claims that its optical modulator clocks in at 10-GHz, tens times that of Intel's optical modulator researchers talked about last year. Since the optical module exists as its own entity, it will require a standard CMOS processes to integrate the optical waveguides. Luxtera has worked closely with Freescale Semiconductor to develop this technology."
Electrons ARE light particles.
Alito: A vote for Alito is a punch in the eye to put that bitch back in her place!
My photons are faster than yours!
And who gets to use these? Are these like only special coprocessors for million-dollar supercomputers? Are they going to be x86-compatible? MIPS compatible? What?
10ghz
time is a perception of a being's consciousness
time is your 6th sense, the wierd ones are 7+
to use light for your processor.
Imagine Intel chips with this technology. Now instead of heating your whole room - you have an extremely bright night light. "Sleep" or "hibernate" will have a new meaning when you use it to turn off the main light in your room.
Now that they have pushed how important Mhz is, will the release of this processor have people buying it up instead?
My thoughts are that it won't matter since the average consumer recognizes the Intel name and have no idea who (*gasp*) AMD or Luxtera are.
It's high bandwidth (10Gbit/sec) small scale (130nm) modulation from CMOS to optical. This is not "processing" in the sense of optical logic.
Lurking at the bottom of the gravity well, getting old
I see the light!
They'll just release the exact same chip as this guy in a year, only call it "Pentium 4 with photon extensions" and pretend they invented it.
At first, I thought "Wow! That's like blazing fast speed!" And then I thought "Well, that sure beats having a couple PS3 cell processors hooked up together" And then i read the article... and was promptly disappointed. The 10GHz speed is how fast it can turn electrons into photons, but the chip is still primarily electron-based, so what is the real performance gain? They don't tell you because it probably isn't any yet.
minus the omniture spyware tracking and massive banners
________
Startup Luxtera has announced its plans to enter the CMOS photonics market, anticipating the day when microprocessors will transmit information via light, not electrons.
The company claims that its optical modulator for transforming electrons into photons runs at 10-GHz, ten times the speed of an optical modulator Intel Corp. researchers began talking about last year. Beginning in mid-2006, Luxtera hopes to enter production of photonic devices using standard CMOS manufacturing processes. ADVERTISEMENT
Although the majority of chip-to-chip communications are conducted using copper-based interconnects, researchers are already looking toward the day when the balance shifts toward optical transmissions, initially for chip-to-chip interfaces between microprocessors, or between a microprocessor and memory device. Fibre optics are a standard component of modern telecommunication infrastructures, and interfaces such as Fibre Channel also use optical fibre interconnects to link up devices.
Although light slows down by some degree when transmitted through an optical medium, shifting to optical-based components is still too expensive than relying solely on copper, even when factoring in the additional power, heat, and crosstalk issues.
"The problem here that we can solve is a matter of bandwidth," said Gabriele Sartori, Luxtera's vice president of marketing and a former advocate for the HyperTransport protocol developed by Advanced Micro Devices.
Part of the relatively high cost of photonics comes from the fact that converting electrons to photons requires an intermediary device, such as the modulator Luxtera is designing. Today, that device exists as a separate module. Intel, Luxtera, and others are trying to integrate the optical waveguides within standard CMOS processes, that can be controlled by the standard voltage swings of a microprocessor.
However, doing so requires that the optical vendor have close ties to a microprocessor manufacturer. At Intel, that's no problem. Luxtera, on the other hand, has worked closely with Freescale Semiconductor to develop the technology. Finding a partner like Freescale is "necessary," Sartori said. "You must walk before you can run."
Freescale has taped out several engineering samples of the optical technology, including a chip, one side of which includes the optical interface built in. The sample chip use a 130-nm SOI process, the same technology used to fabricate the G4 microprocessor. Part of Luxtera's job has been to develop silicon libraries, the files used to design the photonic chips in the same way other libraries serve as the blueprint for making more conventional semiconductors.
The 32-employee startup originally received $7 million funding from Sevin Rosen Funds and August Capital in 2001, followed by an additional $15 million by New Enterprise Associates in 2003. Eli Yablonovitch, a professor at UCLA who developed photoelectronic crystals, sits on the company's board, while Arno Penzias, who won the 1978 Nobel Prize for his work on the Big Bang theory, serves in an advisory role. Other board members include Andy Rappaport of August Capital, which funded Transmeta, among others.
"Moore's law will finally stall out, you can only make electron paths so small and then you hit a physical barrier".
Oh yeah? Well then we'll just stop using ELECTRONS! BOOYA, bloggers!!!
The article isn't about pure photonic processors that working completely on light. These would be used in fiber optic routers where switching between light and electric signals is a waste. Data is already transmitted via light but the modulators used are seperate from the computing logic parts. In this new system, the computing system is still using classic transistors but the optical parts are integrated onto the IC. This is still a far step from pure photonic computers where the "transistors" or logic gates are done purely via light.
--
Want a free iPod?
Or try a free Nintendo DS, GC, PS2, Xbox. (you only need 4 referrals)
Wired article as proof
If you overclock one of these babies and it starts to glow, that isn't necessarily bad...
This makes a lot of sense for an interconnect between chips on a single board.
Transcend Humanity. Please.
IBM is working in this area also . . .
Will be interesting to see a PowerPC with the guts of the VMX unit running at 10Ghz . . .
Anytime you're using light for any purpose-- say, to light a parking lot, or to communicate between links in a processor-- and the light is visible to anyone or anything not explicitly served by that purpose, that's bad. If you're lighting a parking lot, and your lights are visible to planes passing overhead, that's bad, because that means you're paying to light the night sky for no reason. Similarly the chip manufacturers are going to want to make sure the photons stay inside of the microchip; if any is visible outside of the microchip, that means the light is leaking. Every bit of leaking light is wasted power. It would be like if your car was driving down the road and dribbling oil behind it as it went. You don't want your car to waste oil like that. You don't want your computer to waste light and therefore electricity either.
Ah yes, brutal dissappointment. As I read a bit further it looks like this is not optical logic.
pfth.
Does it run Linux?
Also, while I'm here...
Imagine a Beowulf.....
In Soviet Russia photons....
3 PROFT!!
Only old people use photons....
"I'll waste 'em with my crossbow!" ~Bob Herzog, Power Gamer
I just looked around my living room and found that I have nothing electronic made here, save for the front door lock that was made by a company in Ohio.
YOU FAIL IT!
I'll believe it when I see it in action.
READY.
PRINT ""+-0
http://www.forbes.com/forbes/2005/0411/068.html
Interestingly, the 10Ghz figure comes from a measurement made a researcher at Sun Labs, who have been working with Luxtera for more than a year now. The article also talks about what other companies such as Intel and IBM are up to.
Just thought I'd clear up that potential confusion...
Lawrence Person (lawrencepersonh@gmailh.com (remove all "h"s to mail)
http://www.lawrenceperson.com/
Actually this is less dissappointing that I originally thought --
A major problem as CMOS processes get smaller and smaller is wires and wiring. Its really bad at 90nm and it looks like its going to be way worse at 65nm.
Even if optical interconnects can just be used for long intra-unit busses (think L1 cache to fetch/decode unit, and there to integer unit and float unit, etc) we could see great performance gains.
Something like when the upper metal layers in CMOS went to copper a few years ago.
Although the clock on your cpu may be 3 GHz, the speed that you can communicate with the memory and other hardware is considerably less. It is actually quite difficult to get really fast speeds going on a printed circuit board. (My students have been able to get data moving at 6 GHz on a pcb so I'm fully aware that it can be done.) Being able to communicate with the memory at optical speeds could make your computer orders of magnitude faster.
Remember that they are talking about serial communication however. To compare that to a 64 bit wide data bus, divide by 64. 10 GHz / 64 = 150 MHz, not that fast. So, you would need several modulators to actually see a speed improvement. If we look at this as a disruptive technology, it holds promise.
The subject says it all... just hope they don't launch Data in to space (yeah, yeah, he was destroyed in the last movie [sob]).
So now that these non-processing "processors" are available, what has to happen before benefits can appear to people who just go to Fry's and buy things? What has to happen, someone has to license these thingies and put them into a chip or a motherboard bus or something, and then we'll buy the chip or the motherboard bus? How many years can we expect before end-user productization happens?
This is sad... you have a Mac mini because it's cute? I bet you drive a VW bug with a flower vase in it too.
Other groups working on optical interconnects: (incomplete list)
Heriot Watt
Cornell University
IBM Zurich
Delft
UIUC
Intel
Stanford
Sarcasm never works well in text does it?
Sometimes my arms bend back.
This will be like Crusoe. Solid technology basis. Involves significant tradeoffs. Market fails to materialize. Technology goes back on the shelf where if can be "discovered" again in 5 or 10 years.
Maybe someone smarter than I can explain how it all works.
Okay, I am down with light based switching mechanisms and all that. But in my mind, I'm wondering how registers are "storing" information. Light, to my knowledge, cannot be effectively stored. I recall from a couple of years ago someone attempting to make progress in that area but I don't recall hearing that they were successful.
I guess it's time for me to go back to school on this new technology 'cause I *just* don't understand it. Anyone who does understand it care to spit out a few paragraphs to summarize how it works assuming the reader already understands the basics of digital electronics?
At 10 GHz and an index of refraction of 1.5, each 2 centimeters of light pipe adds 1 clock cycle to the latency to the system (2 clock cycles to the round-trip). Put a optically-connected device a foot (30 cm) from the processor and you have 15 clock cycles of data (or a 30 clock cycle response time) just due to the fiber, let alone any in the devices at either end of the fiber-optic pipe.
Its always interesting to see what happens when the relative speeds of processor, memory, and interconnects change.
Two wrongs don't make a right, but three lefts do.
I cant believe it!! That viewsonic computer runing at 10Ghz on amazon.com was real. I'm sure kicking myself for not buying it.
Now when I find a bug in my code I can just reconfigure the photonic matrix and reverse the polarity of the power coupling.
And if that doesn't work I'll try modulating the field harmonics.
This can really save me in a tight situation.
Robert
Bet this
It is a light, portable screen usually circular and supported on a short-term scale, but ultimately, they're just masking the real problem, which can only be solved by the level of thinking that created them.
The average girl would rather have beauty than brains because she knows that the nature in which ramanujan was referred to as "indian math guy" in the sense of optical logic.
That's an old fashioned honest to goodness Troll in the classical sense as opposed to "expresses an opinion I disagree with sense."
Can you see the difference? Smell the quality. It's elegant, and beautiful in its simplicity, while being powerfully provocative.
And look at you being the pendant, all modded up Informative with your response. He's exactly why Slashdot needs a +1 Troll option. And it should award double karma points.
Where's the kaboom? There was supposed to be an earth-shattering kaboom! --Marvin the Martian
They say the first thing to go is your penis. Well, it's either that or your brain. I forget which...
Yes, I'm terrible at it.
BUUURN!!!
this won't affect the resolution of my porn, only its transfer rate.
Sarcasm never works well in text does it?
doesn't the <Sarcasm> tag mean anything to you?
The transistor was invented in America, fuckstick.
I just looked around my living room and found that I have nothing electronic made here, save for the front door lock that was made by a company in Ohio.
Probably not, but why don't you check who had the original patent on whatever you look at? Commodity equipment is meaningless, and has nothing to do with R&D. Who developed the microprocessor? Telephone? TV?
America's research and university system is what has allowed it to dominate a lot of R&D for the last 50 years. Too bad we're stifling innovation at our national labs through bureaucracy and cutting funding to university research. Yay! Maybe you'll be right in 50 years, but you aren't now.
Yeah, but will it play Quake?
"It's difficult to meditate on amphetamines." - Joe Walsh
torpedos, darn it!
You can't blow up stuff with a little processor. Torpedos or nothing!
I could see new PBX systems being made out of this, where a fiber optic line could come into a company (or maybe 2 for redundancy), then running a fiber line to each floor of a building, where it would convert it to digital lines for traditional office phones. I don't know how many channels fiber can carry, but in a 10 Ghz signal, there's room for 100,000? 100 Khz signals.
Combine this in a lot of buildings and all of the 'dark fiber' out there, and this could give VOIP a true run for it's money. Plus it keeps all the VOIP traffic off the internet.
Sounds like the next evolution past T1/E1....
Then again, with that kind of speed it could still be voip, but at CD quality sound. Though there are still potential legal differences between VOIP and telephony.
Innovations like this keep the future interesting.
is definitely not listed by W3C. So you shouldn't be using it at all. How many times do you people have to be told, avoid non standard tags!
No. These are not "processors" of any sort. It is a new way to modulate signal between CMOS and optical at high frequency and small scale. It may provide faster bus speeds, assuming the reality matches the funding hype.
I suppose then that putting them as the data bus to memory would be the best first thing to do. Imagine being able to read memory at register reading speed, that would be great even if you keep your same Pentium IV processor.
I'm beginning to suspect that the Slashdot editors don't always read the news articles before posting them.
Finaly, the end of heatsink age.
I wonder how to overclock one of these processor...
Honestly, when we've already come so close to getting teleportation to work in real-life scenarios, why even bother with light? The way I see it, we're using a defunct technology already, and we'd be better off just skipping it and going directly to teleportation. Face it, if it's not instantaneous, it's too slow!
what are they going to call this baby? Goethe?
The face of 'evil' is always the face of total need
Just think about the great achievements announced over the past 12 months regarding prototype devices. No. Wait. There haven't been any. Don't hold your breath - it will be years and years before anything like this "enters production".
Sorry, it's just BS.
If you read the article carefully (which is laced with marketing hype and was obviously written by someone only passingly familiar with the technologies involved), you will see that nobody's promising optical cpu's in 2006. In anticipation of future optical chips and other technologies, Intel has begun developing one of the stepping stones toward this technological era, which is an optical/electrical gateway of sorts which can be built on a standard electrical chip to allow it to interface with optical components. Think a modern cpu, with some low level optical/eletrical interface on the edge of it so that a row of optical "pins" can stick out one side in addition to the normal electrical pins on the bottom.
This little startup company is working on the same thing, and hopes to have it out soon. Their marketing article is trying to build hype so they can get more cash. Nobody will be selling anyone an all-optical cpu in 2006 (or 2007, or 2008, etc).
11*43+456^2
The tech support calls for this will trump all: Tech Answer 1: Data loss? Ma'am, it says clearly in the instructions that this device is not to be used near any singularity of any kind. It's been known to warp and bend results. Tech Anser 2: Sir, the machine is acting slowly? Are you by chance going 299,792,458 m/s? That drops performance to 286 levels. What's that? you're running BSD? So why are you complaining? Plenty of horsepower for that.
"It's difficult to meditate on amphetamines." - Joe Walsh
From the Article:
The company claims that its optical modulator for transforming electrons into photons runs at 10-GHz
I may not have a Nobel Prize, but I do have a Ph.D. in physics. Electrons do not tranform into photons. They may produce photons, but not turn into them.
I see these articles that claim the creation of optical processors. But read the article, and all the researchers have to do is add a silicon processor and BOOM, we have an optical processor. It's not that easy.
I remember the researcher who created an optical computer that was the size of a room. Why is this? Electrons are small. They bend around corners. They stay put. They move when you want them to. Photons do not bend well around small corners, do not support CMOS-like circuits and generally fail at most tasks of that versatile, tiny doer of great deeds, the electron.
As usual, it's just an optical modulator. Boring old modulator.
It may take a few nanoseconds for the light to bounce around, but that light can be modulated at extremely high rates (that electrical wires cannot). Managing latency is a well understood problem, generally solved by using speculation, buffering, etc..
The fact is, if these parts are running at 10ghz, you will have 10ghz connections between connected parts (with a few nanoseconds of latency, which is mostly irrelevant).
Bandwidth is a measure of frequency and number of communication channels. This advancement does indeed provide more bandwidth, mostly because it can be clocked higher. All computer configurations could see substantial benefits because current electrical designs have highly limited bus speeds (it is not signal propagation that matters, but signal modulation speed "frequency").
Again, signal propagation speed is mostly irrelevant. Signal modulation speed is what is important. Latency != Frequency.
The reason that it can be true that 1+1 > 2 is that very peculiar nonzero value of the + operator
One of the interesting things about light is that it is non-interfering. In other words, it should be possible to lay down multiple signals in a single fiber instead of having to bundle multiple fibers together. There have been many advancements in the communications industry around this topic that could be relevant here.
By loosing the restrictions imposed by the PCB, it should also be possible to have much more ingenious designs. What this tech could do for SMP alone is staggering in it's implications (A Rack / cluster that acts as a tightly coupled SMP solution?).
The mind boggles.
One question I do have is whether they will be able to use this on-die as well as off. Can they replace electrical wiring on the chip itself with photonic pathways?
The reason that it can be true that 1+1 > 2 is that very peculiar nonzero value of the + operator
Photon processors are boring; what we want are photon torpedos for people who submit pointless comments.
Maybe now I'll be able to crank all the settings up in Doom 3.
Is it just me or is this a really badly constructed sentence? It changes subject halfway through (from the speed of light in optical medium to the cost of copper).
Oolite: Elite-like game. For Mac, Linux and Windows
1st April yet, but you'd better expect these jokes.
The article is full of hints which show us it is one... AMD developed HT? WTF?! LOL
All we need now is for some Taiwanese guy to come along, increase the speed of light, and con us into buying his super-fast version. But it won't work this time... we're on to him...
We're geeks... We're the sorcerers of the modern-day world. --
As far as I can tell from the mediocre article, all they've done is gotten a tiny LED to flash at a rate of 10 giga flashes per second. I guess you could call that a "modulator". But it's just a fast LED. And they havent explained how it's going to be a economical and compact way to shuttle data around.
I click this topic and not a single mention of the word "Torpedo". Geeks must be losing their luster.
Hmmm...I wonder how "big" one of these interconnects is. Currently chips have pins that are visible to the human eye, and even solderable by the human hand. If it were possible to have optical "pinouts" that were really small, you could decrease the size of a chip package/circuit board. Of course, I suppose there would still have to be pins for power, but ya' know.
And since an optical interconnect wouldn't need solder, these chips would need a completely different process for connection to a circuit board. No solder means that the chip could be affixed at a much lower temperature. Production lines already use machines to monitor circuit board assemblies to make sure all chips are lined up, but with optical interconnect, they'd probably have to be even more precise.
they have this already: DLP
Hey, we all came from somewhere else. It's the country, not the ethnicity that's making the difference.
Actually, no. America's vast capital reserves and efficient capital markets have allowed it to dominate R&D for the last 50 years. Russia and China have produced a huge amount of innovative and inventive research, but lack the virtuous, steal, innovate, capitalise, commercialise, rinse, repeat cycle that America excels (excelled?) at.
Hand in glove. America's university and research system is the best by far because it has resources and freedom. China hasn't done much (though it's improving very quickly), and Russia's done practically nothing since the fall.
An inordinate amount of the difference making inventions in the last 50 occurred in America. Often by people not born here. That's why the university system works so well - because something like 48 of the top 50 research universities are in America, we get to steal and often keep the best talent. We aren't winning because we're smarter, as a whole.
The rest of the world subsidises US growth by propping up the currency, having it's best brains and talent immigrate there and bending over when it's technology is stolen and commercialised by American companies. In return for this subsidy, the rest of the world recieves access to the world's largest export (or import) market for it's commodities, tourism and specialist products.
The US currency didn't get that way because people decided it would be a nice thing to do. It's a result of all the other things that made it stable and worth something. Foreigners come to the US for college especially in science/tech because they're the best by far in terms of resources, human and capital. Many of those people stay.
That won't matter as long as the US capital markets remain transparent and effecient. The current trend of poli-corporate-criminals indicates a trend towards corruption and opacity of those markets.
That'll work for 10 years but your view is too static. Innovation drives success. Capital's important but if you don't keep ahead technologically, you end up in businesses with no margins and suddenly your capital becomes worthless. Consider if we hadn't invested in tech and remained an agrarian society. The best financial system wouldn't mean a thing because we wouldn't have anything worth investing in.
Everybody calls them "electrical" computers. So once we move on to light, will people change the name? I always thought it was a stupid distinction, personally. I mean, you would call mechanical engineering "metal engineering".
When newer processor technologies are developed, they're almost always deligated to server processors before they trickle down to desktop processors. (Of course, there are exceptions: MMX and its spawn, etc). Should've stuck to your guns :) MMX is vector processing technology, which started in supercomputers like Crays AFAIK.