In 2003, everything old was new again, and new things had become old. This wasn't good for startups, especially flashy, pie-in-the-sky startups promising "revolutionary" products.

Elaborate new technologies that required ripping out and replacing old equipment were rejected outright. Carriers' interest was limited to anything that would enable them to sweat the old network for another 12 months.

As if that weren't bad enough, venture capitalists continued to hide in caves, so even those startups that had promising developments had trouble finding the cash to continue.

Combine hesitant operators with coy VCs, and you get a miserable environment for startups. It's easy to see why network equipment startups selling flashy new kit in 2003 had such a tough year.

Here's a rundown of some of the startup burnouts we reported this year:

No. 10 Crescent NetworksThree months after it declared that it absolutely wasn't closing its doors, this edge router company did indeed cease operations (see Crescent is Waning).

Crescent's routers fared well in tests with BTexact Technologies, the testing arm of BT plc (London: BGC), but it was unable to turn these early trials into revenue-generating sales and shut its doors in January.

After raising $66 million, the assets of the company were sold to British network equipment manufacturer Marconi Corp. plc (Nasdaq: MRCIY; London: MONI) for "tens of thousands of pounds," according to a source familiar with the deal. Did somebody say FIRE SALE?

No. 9 Silicon Access NetworksThis network processor company's chips were up sometime in October, but it was hard to tell exactly when Silicon Access shut its doors, as it disconnected the phone lines before Light Reading had the chance to report a proper postmortem (see For the Masses).

Nevertheless, our stories leading up to its disappearance shed some light on what happened.

Silicon Access made a big fuss about a deal it claimed to have signed with Huawei Technologies Co. Ltd. in February. The Chinese router company was said to be using its network processor, address processor, classifier, and accounting device for 10-Gbit/s port speeds (see Silicon Access Nabs Huawei).

But for months after this announcement, Silicon Access's competitors insisted the deal never materialized (see Huawei Chip Deal: Who's Got It?).

Silicon Access officials stood by their story. Chief operating officer Rex Naden wouldn't elaborate much, citing Huawei's reluctance to publicly disclose its plans, but at the time he said Silicon Access "absolutely" did sign a contract with Huawei, claiming revenues were on the way.

These revenues never amounted to much, clearly, as Silicon Access has not been heard from since. The company had raised a total of $124 million in funding.

No. 8 Corona NetworksIP edge router startup Corona was in business for six years before it eventually kicked the bucket in August (see Corona Networks Disappears).

Corona's end came as a bit of a surprise, as just a few months earlier it had squeezed $8 million in extra funding out of its VCs and clinched a deal with Alcatel SA (NYSE: ALA; Paris: CGEP:PA). The French equipment maker signed a contract with Corona to produce a broadband remote access server (B-RAS) for its 7301 DSLAM (see Corona Gets a Boost).

Then, mysteriously, Corona was spotted in talks with Zhone Technologies Inc. (Nasdaq: ZHNE) around the same time, but this conversation cooled, and Zhone went off looking at other IP edge router/subscriber management services companies.

Did Corona blow it by hedging its bets on both of these deals? Whatever happened, it's time was up. It had raised a total of $78 million in funding.

No. 7 Tenor Networks Tenor Networks raised $120 million to build a giant IP/MPLS switch for service provider core networks. The problem? Cisco Systems Inc. (Nasdaq: CSCO) and Juniper Networks Inc. (Nasdaq: JNPR) had already started building MPLS functionality into their existing routers – thus rendering Tenor's box obsolete before it even hit the shelves. Bummer!

In addition, carriers were looking for equipment that could help them gradually migrate existing ATM services onto an MPLS backbone. The Tenor box required carriers to rip out old kit, which really doesn't go over well in hard times when no one has two beans to rub together.

Tenor tried to change with the times, but it ended up being behind the market with its next product and eventually pulled the plug in February (see Tenor Goes Silent).

No. 6 Metro-Optix Metro-Optix, one of a bunch of startups targeting the next-gen metro-optical networking sector, built and sold its multiservice provisioning system to 15 customers but still failed to keep its head above water (see More Cuts Coming).

As the need for capacity continued to wane, so did many of the startups that were the heroes of optical networking a couple of years ago.

Metro-Optix, which raised a whopping $136 million, ended up auctioning off its assets to Xtera Communications Inc. in a fire sale deal in August, the details of which were never disclosed. One Light Reading reader had this to say about the achievement:

Oddly enough, Xtera, flush with $30 million in funding in August, is still hanging in there (see Xtera Scores Surprise $30M).

No. 5 CeyYa! Ceyba Alas, the same fate did not await Ceyba (formally Solinet Systems Inc.), another optical long-haul startup like Xtera (see Ceyba Shuts Down).

In August Ceyba was abandoned by one of its VCs, which pulled out of a round that was expected to happen later this year. The move set the cat among the pigeons as the rest of board also bailed on the plan, deciding that it was too risky to keep Ceyba going. [Ed. note: And a thousand lemmings can't be wrong!].

Ceyba had at least two U.S. carrier customers and a war chest of $93 million in funding, but this still wasn't enough (see Ceyba Rattling in Ottawa).

Startups like Ceyba were hit hardest by the downturn, because they specialized in products for core network capacity, where the most extravagant excesses of the boom era were focused. What's more, most next-gen equipment for core deployment calls for carriers to commit to a new network architecture that's different from their current, Sonet-based gear. For the majority of 2003, carriers balked at any such changes, as they drop all but the most urgently needed network upgrades.

No. 4 Innovance NetworksAnother long-haul letdown, Innovance shut its doors in December after failing to secure additional funding, according to several Canadian news reports (see Innovance CEO: Layoff a 'Rebalance', Company Makeover).

Innovance's plan was to provide end-to-end optical transport for carriers, incorporating a kind of "wavelength-on-demand" style of provisioning. With capex spending still frozen and excess capacity still a major problem, Innovance went in-a-trance and never came out again.

The company employed more than 310 employees in February 2002 and had raised more than $130 million in funding since it opened for business in May 2000.

No. 3 Network Photonics Fancy all-optical switching gear featuring tilting mirrors and prisms that split light were all the rage in 2000 when Network Photonics raised a staggering $106.5 million to build some.

Unfortunately, these prisms are now tripping the light fantastic on eBay for $10.99 a pop, as such experimental technology got dropped like a lead balloon during the downturn.

Seriously, though, if you're talking about splitting the atom, or whatever Network Photonics was doing, when everyone else is talking about maintaining the tin cans and string, what do you expect?

The company did attempt to regroup and do something else, but it failed abysmally and eventually shut its doors in April (see Network Photonics Scales Back, Network Photonics Shuts Down).

No. 2 OMMOMM, once the leader of the pack in the all-optical switching game, faced the same gloomy fate as Network Photonics – only its demise dragged on even longer, as most folk expected the company to pull through (see OMM: The End Is Near).The all-optical subsystem vendor had been trying to land more funding since mid 2002, and given that it was shipping product and had paying customers, this didn't seem impossible.

OMM counted Ciena Corp. (Nasdaq: CIEN) and Siemens AG (NYSE: SI; Frankfurt: SIE) among its customers, but it turned out that most of its work was still going into lab trials rather than live networks, which doesn't pay the bills (see OMM-inous News).

OMM decided to take a stab at 3-D MEMS (more light-splitting prisms and tilting mirrors), which no doubt gave the VCs the willies and contributed to their eventual decision to pull out. OMM closed its doors in March, laying off 85 employees. It had raised close to $100 million (see OMM Closes Its Doors).

OMM's demise spelled the end of all-optical switching in 2003. The question now is: Will this sector ever come back?

No. 1 PhotonExAt the top of the pile of companies that hit bottom in 2003 is PhotonEx, which filed for Chapter 11 bankruptcy protection in November (see PhotonEx Falls Into 40Gig Hole).

PhotonEx, founded in 1999, claimed to be selling "the world's only commercially-available, field-proven," 40-Gbit/s, long-haul DWDM systems. To do so, it raised an astounding $178 million in three financing rounds.

"It was a classic case of a company with technology too advanced for what carriers wanted," says Scott Clavenna, chief analyst at Heavy Reading.

Unfortunately for PhotonEx, carrier budgets didn't allow for 40-Gbit/s systems nearly as quickly as the company had hoped. Sources say shortly after the company failed to get any part of the U.S. government's Global Information Grid Bandwidth Expansion (GIG-BE) business, its managers decided to wind down operations.

For anyone who's counting, the total amount of funding raised by these 10 companies was $1.13 billion. Which would almost cover the Light Reading staff's Christmas bonus!

— The Staff, Light Reading