Unstealthed: Was Solexel a Bad VC Bet?

At the end of last week Eric Wesoff and I reported on Solexel, a stealth solar company backed by Kleiner Perkins and Technology Partners. Those reports can be found here, here, and here. A few more things have unfolded since then.

First, it appears the Kleiners Perkins IT department has finally gotten their act together (or at least paged through Back End Maintenance for Dummies), and removed all mentions of Solexel from their website after we called them out on it last week.

After the jump, some major doubts about Solexel’s non-standard MEMS technology. Did Kleiner Perkins back the wrong nanohorse?

An anonymous tipster from deep inside the nanotech industry tells me Solexel’s technology “would be tough to scale to some meter-size scale if the goal is large format - that may require some etch tool development, and a tool partner that supports it.”  Also, “maintaing the Si quality throughout the drastic conformity/geometry changes of the wells may be new territory,” but Solexel “can’t leave the wells exposed to air, so the wells need to be filled with an optically efficient solid medium, or the entire thing needs to be sufficiently encapsulated. Both of those things affect coupling efficiency.”

Bo Varga, nanotech expert and Managing Director of Silicon Valley Nano Ventures, puts it into plain English: the technology will not “impact the markets in less than 5 years and move markets in less than 10 years.”

So, is this a really good example of a VC firm getting over-excited, betting on a non-standard technology still in the R&D phase, with little hope of commercialization anytime in the near future?  The bottom line with Solexel’s non-standard MEMS technology, according to one analyst, “is that there is an enormous technical risk with this stuff and it is really a place for DOE, NSF, maybe angel funding and no place for VCs. It is indicative of falling in love with technology and forgetting about costs and value propositions.”

The problem with non-standard MEMS processes is that they require “their own MEMS fab,” which the VC “has to care and feed the foundry and all the design, process, and sustaining engineering that that entails.” VC’s are traditionally loathe to finance large, capital expenditures like fabs and factories. According to that same analyst, most recent non-standard MEMS startups “all died quick and sometimes not so quick deaths. The survivors with MEMS technology proved to be conservative incumbent firms.”

Here’s hoping John Denniston knows what he got himself into.