The 1st Commercial-Grade All-Optical Switch?

joabj writes: "Today's Washington Post trumpeted the delivery of the first-ever commercial-grade optic-switch, by Corvis Corp. I'm not sure how commercially viable a switch the size of a refrigerator can be, though a commercial-grade all-photonic switch is indeed, as the WP points out, a "holy grail" of the optics industry--the missing piece for the 40-80 Gbps throughput for fiber optic cable that the industry is so hungering for these days. (Now, photonic signals must be converted into electrical signals to be switched, then back into light again -- slowing transmission speeds considerably.) Unfortunately neither the article (Free reg. req.) nor Corvis' own Web site explains what the winning technology was. There are a lot of competing ideas out there on how to switch optical signals. Is it MEMs? Liquid crystals? Curious minds want to know ..."

Initially, this sounds a lot like the optical switch announced by Agilent a little while ago, but with the bonus of being ready "within a few weeks" from workaday life. (Does that mean the equipment reported as delivered is still being installed or set up?) [Updated 11th Oct. 2000 4:05 GMT by timothy:] As a few people have pointed out in comments, that free registration is pretty painless ... since it's not necessary. Better than the other way around! ;)

Comments

Probably Microscopic Mirros

[tylerh]

The article says:

Lucent...said it delivered a similar all-optical technology...last month to Global Crossing

Lucent says here they are using microscopic mirros, so corvis probably is too.

Re:Probably Microscopic Mirros

[Puk]

The biggest problem with all-optical switches isn't the routing of the optical signals. They have several technologies, including MEMS-controlled mirrors, for doing this right now. Optical cross-connects are commercially available (though expensive) products.

The difficulty with optical switches is in determining _where_ to switch the data without converting it to the electrical domain first. Plenty of schemes have been proposed where the data goes O->E to read the headers and decide where to switch it, and then optically switches the packet farther along the path (where the actualy signal of the packet never left the optical domain), but no one has come up with a switch that does it _all_ in the optical domain. Admittedly, few if any of these "mostly-optical" systems have been built.

I'm not sure what Lucent and Corvis have done. It sounds to me like they have probably implemented one of these systems. They claim it's "all-optical", but I doubt that they have built this without at least the headers being converted to the electrical domain. Even so, it's an impressive achievement to have made a commercially viable product which does that much. It remains to be seen which way will end up being more cost-effective in the long run. After all, that's what matters in the industry. :)

-Puk (Yes I'm Probably Wrong)

Commercial switches vs. prototypes

[joabj]

There are a lot of all-optical switches out there that are being worked on. In addition to Agilent, I can think of recent announcements by Chorum, SpectraSwitch, CoreTek, Nortel, and Tellium's optoelectronic hybrid model. . . .

Personally, I take announcements of prototypes such as those above with a grain of salt. A prototype is still further away from production than most people realize. You can pretty much do anything you want within the limits of physics, given a big enough R&D budget (and snazzy enough PR department to tout your work). But for your prototype to A: work *all* the time in the field, B: and for a competitive price, and C: to be compatable with existing standards, are the hurdles that kills 99.5% of all new technologies. . . joab

Data Switch vs Telecom Switch

[Gumby]

One big issue with these optical switches is that they don't switch packets, they make persistent connections like a phone call. So optical data passes through cleanly, to ONE place. So this doesn't accomplish the function of a network switch at all. They're electronically controlled, so it's a lot better than manually re-routing a fiber patch panel, but that's all they do.

Some real information....

[Doco]

First off go to LightReading.com to get some general information on this technology arena.

Corvis's switch is an all-optical switch. No eletrical regeneration involved. What it is switching is wavelengths of light. NOT packets. So, you can take a wavelength of light from one fiber and send it out another. This allows you to set up 2.5Gbps (OC-48) circuits quickly.

What technology are they using - don't won't say. Almost all other people in this arena are basically splitting the dense-wave-division-multiplexed (DWDM) circuits into their individual wavelengths. Then routing them through micro-mirrors. The micro-mirrors allow you to connect any two fibers together optically. Then, the outputs from this are re-combined, optically amplified and transmitted.

This isn't for sure what Corvis is doing - but I would bet money that this is basically what they are doing.

One problem with this is that you can't have two circuits on a fiber using the same wavelength of light. So, you would need something that shifts the wavelength of light being used. Nobody that I know of has a commercial product to do this.

Press blurb about this particular thing is available in a Light Reading article.

A couple of weeks ago, Corvis announced that they had revenue - from this shipment of course.

One more link - Some hints to what technology they are using.

Stock price

[Anonymous Coward]

This is just pure stock ramping. I wish it was true as I have some friends at Corvis, unfortunatly, the switch is a long way off. Management wanted to release this PR after a concept trial that was sucessful, this switch is at least 5 months way from market + it's NOT all optical as headers are converted to good old fashioned electrial signals, the switching path is determined and then the packet is switched via MEMS tech.
Lucent is the closest to an all optical switch followed (a fair way back) by an English company call goouch and housego (I think the spelling is slightly off there).