x
Optical components

Xanoptix Teams Up With Tyco

The future of startup Xanoptix Inc. is looking considerably brighter in light of a strategic alliance with Tyco Electronics, due to be announced later today (see Xanoptix Signs With Tyco).

Under the terms of the agreement, Tyco will resell Xanoptix's XTM product line of parallel optical modules under its own name. It will also buy the guts of the optical modules from Xanoptix and package them itself.

This is quite a coup for Xanoptix, says the the company's CEO, Jim Norrod. "It gives us tremendous revenue opportunity next year," he crows. In fact, on the basis of orders already received, and the agreement with Tyco, he reckons Xanoptix is on track to break even at the end of 2003.

For starters, Tyco has agreed to buy a minimum amount of components from Xanoptix over the next two years. Norrod declined to reveal the exact amount, but did note that it was worth several million dollars next year.

This amount of revenue is unlikely to come from massively parallel backplanes, an application sometimes cited for Xanoptix's modules, bearing in mind that the market for monster, distributed-architecture core routers is in the doldrums at the moment (see Switching Silicon Goes Scaleable).

Still, the reseller agreement with Tyco could give Xanoptix a route into a much broader market. Tyco has "a sales channel all over the world," Norrod notes, adding that it's something startups can't match.

Tyco's backing also removes the stigma of buying from a startup, he adds. Customers that would be wary of buying from a startup would be more likely to buy from an established vendor like Tyco. Another plus point is that Tyco already has the right customer connections. It currently sells 1x12 parallel optical modules under the AMP brand to a number of markets, not just telecom. AMP was a company Tyco acquired for stock worth $11.3 billion back in early 1999. The deal with Xanoptix will allow Tyco to extend its product range to offer 12-, 16-, 32-, 48-, and 72-channel interconnects with aggregate capacities up to 245 Gbit/s (see Xanoptix Lands Cash, Launches Product).

The reason that Tyco is so enthusiastic about the deal, says Norrod, is that it allows the connector vendor to get its hands on the secret sauce inside Xanoptix's modules. Xanoptix calls its technology "optically-enabled integrated circuits," which are chips with both lasers and electronics integrated on the same piece of semiconductor.

"Tyco Electronics believes advanced optically-enabled integrated circuit technology will have a significant impact in optics over the next decade," concurs Kevin Rock, VP of Tyco Electronics’ communications, computer, and consumer electronics products, in a prepared statement. According to Norrod, Tyco has indicated that it would like to develop new configurations of optically-enabled ICs.

Exactly how this partnership will shape Xanoptix's future is not yet clear. The startup is hanging on to its own sales team and packaging facility for the time being, but indicates that this could change. "We could do some interesting things from a manufacturing standpoint and for efficiency," says Norrod.

Key competitors in the parallel optical module market are Agilent Technologies Inc. (NYSE: A), and Infineon Technologies AG (NYSE/Frankfurt: IFX). Startups in the market include Alvesta Inc., AraLight Inc., and TeraConnect Inc.

— Pauline Rigby, Senior Editor, Light Reading
www.lightreading.com
Page 1 / 2   >   >>
lightwate 12/4/2012 | 9:32:58 PM
re: Xanoptix Teams Up With Tyco I don't mean to be an obnoxious cynic, but I can't see the logic in developing such a high speed interconnect in today's market!

By the time anybody is ready to buy they'll be coming down the cost curve on 40G serial solutions! What do they think they're offering? Why should I build this into my future plans?

Is anybody connecting boxes or racks together at these rates? I've seen people demonstrating OC-192, but who's actually using it in the field (nobody!)? Who will need it inside of three years? I'm guessing nobody again!

OK, I can see it being good in an optical backplane in a system where you've maybe implemented a perfect shuffle, but that doesn't seem to me to be enough to build a company on.

Enlighten me if anybody is buying these things, but it just seems like technology for technology's sake and refusing to look at the fact that nobody needs a point-to-point interconnect as fast as this right now, and when they do it's going to be at a lot lower level of granularity than Xanoptix is offering.

I hope I am wrong because I'm getting really depressed watching all these firms collapse and it stinks to see another one getting teed up!
fiberlight 12/4/2012 | 9:32:53 PM
re: Xanoptix Teams Up With Tyco "Technology for technology's sake" is correct. Xanoptix never intended to actually produce anything. The whole game was to "put some lipstick on that [technology] pig" and sell, sell sell. The company that is.
headlight 12/4/2012 | 9:32:44 PM
re: Xanoptix Teams Up With Tyco So what's in this for Tyco? Why would they commit for millions if there is no market? Are we missing anything here, is Tyco so badly informed about this market.

They are supposed to understand this market at least a little bit since they had the AMP 1x12 module (done by Mitel and for what I know never was a big cash maker).

So do they see things that we don't see?
lightwate 12/4/2012 | 9:32:37 PM
re: Xanoptix Teams Up With Tyco >>So what's in this for Tyco? Why would they >>commit for millions if there is no market? Are >>we missing anything here, is Tyco so badly >>informed about this market.

>>They are supposed to understand this market at >>least a little bit since they had the AMP 1x12 >>module (done by Mitel and for what I know never >>was a big cash maker).

>>So do they see things that we don't see?

I guess they must - I don't think companies are stupid or anything like that, but everybody works with different assumptions.

Looking at their web site they seem to be limiting their reach to 1km. I don't see this being good for anything except an intra-office connection. But don't designers universally prefer serial to parallel for that, if the costs are on par?

And if they fell on their faces with the Mitel stuff like you seem to imply happened, what might be different? Maybe they're planning on this for 5 years out?
BlueWater66 12/4/2012 | 9:32:19 PM
re: Xanoptix Teams Up With Tyco Tyco... They're wacky. They've fired or lost every serious active optical component guy they've ever had. As I understand it, all of their actives that are sold today are actually OEM'd for E2O (a start-up). At least that's the rumor. They have great connector technology, but they are absolute idiots when it comes to anything active. Xanoptix is mostly smoke and mirrors, so a partnership with Tyco makes a lot of sense. They can "group think" each other.

I'm supprised to hear the article mention Teraconnect and Aralight as competitors. I don't think either is doing well.
headlight 12/4/2012 | 9:32:12 PM
re: Xanoptix Teams Up With Tyco Its clear this is a niche market, with perhaps some perspective in the years to come. I'm still amazed to see Tyco signing up for volumes for 2003. Was their management sleeping when this was decided?

In the same market we have recently seen Gore step-out with a product which was already in production phase and had multiple source in place. Apparently it was too tough to make money and thus Gore decided to concentrate on core high speed electrical markets.

Conclusion is that Xanoptix had a good reason to be pleased with the deal. Hope for them they will quickly see the stock orders from Tyco rolling in.
lightwate 12/4/2012 | 9:32:01 PM
re: Xanoptix Teams Up With Tyco >>I'm supprised to hear the article mention
>>Teraconnect and Aralight as competitors.
>>I don't think either is doing well.

Are any of these parallel optical component people doing well, even relative to optics in general? I don't think so. It seems to me that they made a big assumption that you would need to aggregate traffic up from the metro to the long haul at 10G very early in the game. So an immediately cost-effective solution was needed. They always knew that long terms they'd get passed by serial optics in price, and they'd be preferable to parallel for designers. Actually, I don't think designers really think too much about optical connections between equipment - they choose cheap and are done with it.

Anyway, without a need to add a lot of bandwidth these solutions are stuck in a holding pattern. Anybody thinking parallel optical can be a company is totally missing the boat.

Just my $.02
beamryder 12/4/2012 | 9:31:46 PM
re: Xanoptix Teams Up With Tyco Isn't the SAN market the opportunity that Xanoptix and Amp targeting? It sounds like this segment of opto is still healthy and IBM seems to be making a big splash in it. Norrod is an old IBM exec. He can probably hook Xanoptix up and drag AMp along with them.

Just a thought
headlight 12/4/2012 | 9:31:36 PM
re: Xanoptix Teams Up With Tyco I guess you're right beamryder. But .... SAN seems to be the straw that all want to hang on. Without any doubt SAN is an interesting market. For what I know Infiniband is the technology of choice for SAN. Parallel TRX at 2.5G/line could be the driver for 12x optical Infiniband cabling. Note that these are ONLY used for the longer links. The majority of links in a SAN environment will be shorter and made of copper. The Xanoptix modules meet the criteria for Infiniband and thus this will be an important target segment for Tyco.
lightwate 12/4/2012 | 9:31:31 PM
re: Xanoptix Teams Up With Tyco >> For what I know Infiniband is the technology of
>> choice for SAN. Parallel TRX at 2.5G/line
>> could be the driver for 12x optical Infiniband
>> cabling. Note that these are ONLY used for the
>> longer links. The majority of links in a SAN
>> environment will be shorter and made of copper.

SAN sounds like a logical place to look for growth, present troubles of EMC notwithstanding. I don't know much about SAN architecture, but would you be better off with lots of low-bandwidth links going of in single directions as opposed to a large pipe from point A to point B? Wouldn't you have to distribute those parallel links at some point, which would be more of a nightmare than just having lots of low speed lines in the switch fabric?
Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE