CodeStream Goes Under

RICHARDSON, Texas -- Optical systems maker CodeStream Technologies Corp., the lone subsidiary of CodeStream Holdings Inc., has closed its doors. At the end of April the company shut down, fired 39 employees, and filed for Chapter 7 bankruptcy protection in Dallas County, Texas.

A startup running out of money is a familiar story in these taxing times. But Codestream had both quirky technology and origins -- and its history reinforces the lesson that there's not much point trying to challenge mainstream technology (in this case DWDM) without adequate financing.

Gordon Werner, Codestream's CEO, says the company's technology -- an adaptation of the technique that's already used to increase the capacity of mobile phone systems, called optical code-division multiple access (OCDMA) -- was the kind of thing that venture capitalists usually seek out (see Metro DWDM: Not the Only Game in Town). It was a new, innovative approach to increasing the amount of data that metropolitan networks can carry, while at the same time making it easy for service providers to turn up and shut down bandwidth connections.

There were, however, several things about CodeStream that weren't attractive. For one thing, CodeStream had a rather confusing lineage. One of its SEC filings from last year states the company was originally incorporated "for the sole purpose of raising capital and then seeking out, investigating, and acquiring any suitable assets or business without regard to any specific business or industry." This describes what's known as a shell company, an outfit that sits around on the public market, doing not much of anything, waiting for an opportunity to come along.

In 1996, the company, then called Commercial Technologies, was formed to work on technology that had come from a government defense and aerospace contractor. In 1999, the company merged with RDL Photonic Integrated Chip Corp., a company that owned some OCDMA technology. In 2000, the company merged with Bud Financial Group, a Texas-based holding company, and received an infusion of $7.5 million in funding from venture capitalists, including SpaceVest, Kline Hawkes & Co., and others.

By November 2000, CodeStream was seeking a $50 million funding round in order to get its product to market. It only netted about $1 million. On March 16, 2001, the company filed to delist its stock, which traded over the counter under the symbol "COHO.OB." By April, the funding spigot had been off for too long, and the firm withered away.

Aside from an odd family tree, CodeStream had the curse of every new technology company -- it had to teach the market about the problem it was born to solve. "One of our biggest problems in the market was trying to overcome -- and forgive me for putting it this way -- the ignorance of the analysts and the press about how services are delivered in a metropolitan network," says Werner.

The next natural question might be: Was Codestream marketing a product in search of a problem, or a problem in search of a product? Werner says it was difficult to convince people that DWDM (dense wavelength-division multiplexing), the technology of choice for boosting bandwidth in long-haul networks, wouldn't work as well in the metro because customers change locations so often and their bandwidth needs are all over the map (see Report Forecasts Metro Winners, Losers ). "The customer can't afford to be in one location on Friday and have perfectly good service, only to show up at a new office on Monday where nothing works."

Perhaps Codestream simply set its sights too high by offering to replace metro DWDM, a technology to which many investors and service providers are deeply committed. "You would have had to change everything about your network to reap the full benefits of their system," says Atul Tambe, vice president of engineering at Templex Technology Inc.

Templex, a components firm funded by Intel Corp. (Nasdaq: INTC), is working with OCDMA as well, but it's positioning itself as a firm that makes fiber Bragg gratings for WDM systems, with a technology migration path to OCDMA -- not as a metro DWDM alternative, says Tambe (see Startup Unveils Fiber "Switch").

The fate of CodeStream's technology, which was about 10 months away from field trials with carriers, is unknown. Most likely, a trustee of the courts will sell off the company's intellectual property piece by piece to pay off its creditors.

As for Werner, he's working as a consultant to networking and optical companies, hoping his hindsight can steer someone else right. "[CodeStream's closing] is unfortunate because the technology's benefit would have been huge," he says. "And now, that may never be realized."

- Phil Harvey, Senior Editor, Light Reading
ng_mui 12/4/2012 | 8:15:11 PM
re: CodeStream Goes Under There was a lot of research into optical CDMA in
the late 1980's, and a lot of problems were
uncovered which showed it to be almost completely
useless. This company was heading in completely
the wrong direction!
realguy 12/4/2012 | 8:15:04 PM
re: CodeStream Goes Under CDMA works very well in free space applications such as cellphone. Add some dispersion and non-linearity as is the case with optical fibers and here goes the neighborhood.
Photonboat 12/4/2012 | 8:14:51 PM
re: CodeStream Goes Under There were a lot of papers on OCDMA presented at OFC (that I didn't pay attention to). Many companies are doing research in this area. What are the major problems with this technology if used in short reaches? In a short reach (say < 10km), dispersion isn't that big a problem. By only using one laser for several channels, OCDMA could be useful in the metro.
eewhiz 12/4/2012 | 8:14:42 PM
re: CodeStream Goes Under "CDMA works very well in free space applications such as cellphone. Add some dispersion and non-linearity as is the case with optical fibers and here goes the neighborhood."

Actually CDMA did not work very well initially in "free-space" environments. There is no time dispersion in free-space, but then again none of us live in free space...There is considerable dispersion in a typical mobile channel. Here the channel is changing path loss +/- 20 dB due to Rayleigh fading depths (the higher the vehicle speed, the harder the DSP has to work), and to accomodate adaptive (inband signalling) power control every 1ms, and combat near-far problems of overlaying 50+ Direct sequence orthogonal Walsh encoded carriers onto one channel (64-ary orthogonality was/is the target).
To say that CDMA was cakewalk in the wireless world is not understanding the numerous challenges that had to be solved...AS for non-linearities, the power amplifers are essentially non-linear devices trying to be tamed into linear operation, and the receiver front-ends can be overloaded with cross-modulation, 2nd and 3rd order intermod products (Notice that your cellphone worked in Supercomm with all that interference)
FYI (a scarred warrior who use to work in CDMA)
realguy 12/4/2012 | 8:14:24 PM
re: CodeStream Goes Under Dear eewhiz is right. CDMA was no cakewalk even in free space applications. Lots of innovations by smart engineers were implemented to make it to a commercial success we know today as cellular. But my point was in order to make CDMA a success in optical fiber domain require again a set of equally challenging problems to be solved. Of course 10 km span may not pose a problem but these day's people are talking the possibility of 1000-km metro spans.
jimbeed 12/4/2012 | 7:31:15 PM
re: CodeStream Goes Under eewhiz, I can I contact you to talk about OCDMA in more details
Sign In