Michael Kanellos and Eric Wesoff have covered the trials and travails of Applied Materials' SunFab turnkey production line in recent times in great detail. Bad news has not been hard to come by; from chatter about its possible demise, to the bankruptcy declaration of part-customer Sunfilm, to the refusal of a DOE loan to its Great White Hope (and a company this writer personally likes), Menlo Park-based Signet Solar, to an eyewitness account of CEO Mike Splinter attempting to wash his hands of the product altogether. Observers have been skeptical about Applied's prospects in this market from the get-go, because of the low efficiencies of modules produced and the high cost of the equipment.

Once polysilicon prices fell to the levels they did, anyone with half a brain knew that amorphous silicon PV would have trouble being cost-competitive in the near-term. And yet, like all Big Businesses (for that is what they are), Applied kept confidently denying the validity of any perceived challenges to the technology or the product. Trouble is, they went about doing so with an air that smelled like paranoia. I moderated a Q&A panel in 2009 with a SunFab speaker on it; before the panel began, they organized what was in effect a meeting to approve of my questions. Why be so secretive if you've got nothing to hide? Why would they refuse to acknowledge the rather weak claim that not every single one of their customers would be equally successful? And why would none of their customers ever discuss current costs and prices? One could not help but suspect, as they say in India, that there was something black in the lentil soup. The obvious explanation was the self-realization that SunFab is not a very competitive solution, that its 2010 and 2012 road maps may privately have been acknowledged as either unrealistic or insufficient in a China-made, low-cost polysilicon crystalline silicon PV environment.

Amorphous silicon PV has an almost laughable reputation in some quarters of the solar world, which is not fully deserved. Like many archetypes, its worst examples are held up as representative of its prospects in general. Admittedly, it would be a good general rule to assume that a given a-Si producer is not currently competitive. But how is that different from other thin film? Even here, there are exceptions, and important ones at that. Sharp just installed another 160 MW of tandem-junction capacity in Japan this year, bringing its total a-Si capacity to 320 MW. Mitsubishi and Kaneka ran at utilizations of around 50% last year, which is not bad given how terrible things were. We scoff at "top" efficiencies of 9%, but what difference does efficiency make in a utility-scale market if costs are low enough? This stuff does a hell of a lot better in the desert than a plain-vanilla crystalline module. The fact remains, however, that much work is still to be done, both on costs and efficiencies.

To reduce costs and improve efficiencies, though, you need some serious R&D to develop the equipment. You also need time. So you need a lot of money and a strong long-term vision. You, or your partners, have to be prepared to run at a loss for a significant length of time. This is no different from other thin-film PV. Two things people don't often bring up about First Solar are (A) it took a quarter-century to go from inception to that mystical $1 per watt, almost going out of the business on the way, and (B) it had some serious non-VC (we're talkin' Walton family) bucks behind it. Sharp and CIGS producer Showa Shell Sekiyu (now known as Solar Frontier) are in a similar position.

They say Applied's stock rose the day the original story broke, meaning that most in the market viewed Applied's investment in a-Si as a mistake. I beg to differ. I think the company just went about doing it in the wrong way. PV (especially thin-film PV) manufacturing needs brilliance, objectivity, patience, and commitment. If you don't have these, don't bother playing. Applied should really have been in this business for the long haul, and made clear that its "turnkey" equipment was really anything but, that there would be serious differentiation required for any customer to succeed. Unfortunately, the market turned, and Applied got scared (some will say it also got greedy, that it really should have waited to introduce the equipment until it was demonstrably competitive in a sub-$50/kg polysilicon environment). If the rumors that Applied is indeed pulling the purse strings on Sunfab are true, it signals that what should have been a long-term play was in fact a cheap cash-grab to exploit expensive polysilicon and generous subsidies.