In March 2003, OMM of San Diego, a developer of optical components based on MEMS technology for communications networks, announced its imminent intention to cease operations with the loss of all its remaining 85 employees. The company, whose first commercial products were high-speed optical switches, said that its assets would be assigned to the benefit of its creditors.

OMM's switch modules were intended to act as building blocks for the next generation of fibre optic equipment, while providing carriers with access to technology aiming to decrease capital expenditures, operating costs, power consumption and central office space. Applications for the company's matrix switches in next-generation equipment included: Photonic Cross-Connects (PXC), Wavelength Selective Cross-Connects (WSXC), Wavelength Interchange Cross-Connects (WIXC), Reconfigurable Optical Add-Drop Multiplexers (ROADM) and IP over optical switching/routing.

OMM's President and CEO, Phil Chapman, stated that, as a result of the severe downturn in the worldwide telecommunications industry, "The market timing for volume deployment of our products has been significantly impacted". While OMM had been seeking additional private funding to continue operations, it was unable to raise the total amount deemed necessary to see it through the downturn in the global telecom market. OMM's investors included Sevin Rosen Funds, Atlas Venture, Bessemer Venture Partners, Rho Capital Partners, and Weston Presidio, with strategic investors including Nortel, Siemens, Alcatel, Sycamore Networks, Solectron, Fairchild Semiconductor and Corvis Corp.

The announcement was not entirely unexpected; the company's April 2002 statement that it would lay off 36% of its 136 workforce followed OMM's decision to substantially reduce its new-product development and focus on its existing line of products. At that time, OMM presented as something of a default victory the fact that it was effectively the last man standing following the abandonment of optical switch activities by most of its competitors.

An additional comment from this customer was that MEMS is a good choice for a centrally functioning device within the network since he was sceptical about optical-mechanical devices due to "stitching" - a problem that cannot be simulated as the switch is destroyed in the process.

In competitive terms, the customer said Lynx may fill the gap, though its solution is solid-state, not MEMS, while AO MEMS is also a possible choice, though in any case the board would have to be modified. In summary, he remarked, "anybody picking up OMM's technology would make us happy".

Nevertheless, the fall-out for this particular customer is limited, as OMM's devices are providing functionalty as add/on units. Optical cross-connects are available if an operator requires them. However, as an aside, even a supportive customer of the technology, and of OMM in particular, was dubious about the short term need of any carrier to switch an entire wavelength, asserting that the need would only arise in case of over-capacity, which is clearly not the case at the present time.

However, it seems apparent that while OMM's investors included possible key customers, such as Sycamore, this group was unwilling to provide additional investment and perhaps saw the more prudent path as being that of cutting losses sooner rather than later. While the structure of the deal between OMM and its investors is not known, it was evidently decided that the remaining dollars in the OMM pot would be better deployed by this investment group rather than by OMM in attempting to wait out the downturn. Indeed, even those companies at the more generously endowed end of the scale, such as Corvis, which re-iterated recently that it had some $500 million in cash, cannot say with any certainty that they will see meaningful revenue before burning their way through this substantial cash pile.

Specifically, according to insiders, while OMM was able to raise approximately 60% of what was deemed necessary to run the business to a breakeven point, the company failed in its efforts to secure the remaining 40%. This was despite the fact that new investors were coming on the round, having carried out comprehensive and presumably reasonably positive due diligence.

However, although investors were said to have felt largely confident about OMM's business model, the issues of market timing and time to a revenue ramp, effectively the 'unknown' factors, were seen as the major stumbling blocks. As one correspondant noted, it is clearly impossible to judge a company's business model as sound without regard to the prevailing climate - the two are inextricably linked.

OMM saw itself as offering a unique product with little or no competition, and yet despite - or perhaps because of this - the time to revenue and profitability element was simply too risky. Regardless of the perceived risk, OMM claims that it never lost business on price and says that pricing was not regarded as a reason for the lack of planned deployment of product.

TI's Marketing Manager, New Applications, for the DLP Products division, said, with regard to the OMM closure that "The number of simultaneous challenges became overwhelming. OMM faced a severely depressed market where buying literally stopped. At the same time, various targeted customers changed strategies or went under themselves and there was heavy price pressure, with all solutions becoming ever cheaper as companies struggled to turn sales. OMM faced the problem of trying to iron out the intricacies of MEMS technologies, while driving for adoption of new network architectures, just as the existing network and equipment inventories became sufficient to meet demand for the near-mid term."

Despite this analysis, OMM believes that its customers and carriers, with few exceptions, were comfortable with MEMS technology, in particular the simpler, 2D MEMS that OMM adopted in its designs.

All in all, said the TI spokesman, "That is a very tough hand to play. Ease up a couple of the conditions and who knows what might have happened? They might still have failed in time for technological, architectural, or cost reasons, though equally they might have become viable and successful."

This situation has led to a huge shift in the way network development is viewed. Those of OMM's loss-making potential customers still hanging on with unreconstructed businesses, clearly have become more wary of challenging new technology, and have a greater reluctance to embrace the somewhat radical new optical switching technology. The incumbants grow networks rather differently - incrementally, by adding on to substantial existing legacy networks, with the add-ons remaining compatible not only in terms of equipment but also in terms of opex and network management considerations.

While much of the impetus for this behaviour is explained by the innate conservatism of U.S., or indeed any, incumbent, the fact that in the U.S. incumbents are fined heavily for downtime compounds this instinct, acting as a disincentive on the part of these carriers to 'push the envelope' in terms of signing off on new technology.

In addition, with vendors such as Alcatel, Sycamore Networks and Nortel Networks each facing severe problems of their own, pressure for a modification of their models, product lines, and R&D has become irresistible, leading to a return to 'core competencies'.

OMM's MEMs-based technology, like much new optical technology in general, simply does not fit into this new environment. With purchasing becoming business-driven as opposed to 'speculation-driven', companies understandably have looked for alternative ways to do business. Sycamore, for example, has integrated more optics into its OEO switches - for which the component prices have reduced considerably - and forged a strategic deal with Siemens to add the box to its portfolio rather than develop a new one. The end result is that the Sycamore/Siemens box is seen as offering more valuable features, while apparantly taking up less space and fitting into the kind of networks that exist currently. Accordingly, Sycamore has found itself with better sales access while Siemens has saved on R&D costs.

Despite this, OMM, it is believed, is at least considering a restart, perhaps via a merger, and a growing economy is seen as the most likely facilitator of a re-birth - although, as one insider put it, "Under normal conditions, OMM would still be here, financed and selling product."

Failing that, a sell off of assets and IP would be the likely next path with interested parties already reported to be lining up - indeed there were several rumours at CeBIT in Hannover last week on that subject, with at least one systems provider showing equipment that included OMM's product, - and presumably rather concerned that the source had departed.

While OMM is certainly not unique in failing to succeed - indeed, it is in depressingly good company - one would think that it is unlikely to be the last to go in the MEMs sector. And yet, as Wes Stalcup of TI points out, "It doesn't seem there are many like them left".

Certainly TI believes that carriers have little need for all-optical, in much the same way that a whole network does not need to offer broadband to all its customers from the off, and that carriers will focus on the key areas first and incrementally transition with demand over time. The incumbents know backbones exist, they know they can grow and "these carriers hold most of the cards now".