Then we came across Stion. The company wanted to develop a completely new material that was less expensive and more efficient. We started working together and soon faced a choice: Should Stion go to market with its existing technology, which had reasonable efficiency, or take longer time to market and iterate on R&D to achieve higher efficiency that could compete with silicon? We opted for the latter. We helped Stion recruit a stellar team and focus on iteration.
Stion's focus was to emphasize the cycles of learning as quickly and rapidly as possible. Instead of being wedded to a specific technology, the company looked at a wide variety of semiconductor materials. Cognizant of others' mistakes in rushing to spend capital prematurely on facilities, Stion refrained from building a large facility until the technology was ready. The company acquired several pieces of used equipment along with some new equipment and built a pre-production line with 10MW of capacity. With less than $45M in equity and $10M in debt Stion was able to fund it ongoing R&D, engineering and was able to demonstrate commercial ready product in October of 2009 about three and half years from inception.
Since October 2009, Stion has focused further on yield, reproducibility, quality and cost. In 2010, TSMC teamed up with them in order to get rights to build up to 1 GW of capacity using Stion’s first generation technology. In December 2011, a large Korean consortium led an additional equity round to support U.S. operations and build another facility in Korea. Today, Stion has achieved 90-percent yield on 13-percent circuit efficiency on 65 X 165 cm (nominally 2 X 5 foot) panels at a very advantageous cost-and with far less upfront capital than competitors, and Stion’s first United States manufacturing facility is on-line in Mississippi and producing cells. Stion continues to push the technology further and is already working on its second- and third-generation products. In this case, taking a technical gamble and setting up the company for rapid technical iterations ultimately lowered the risk and cost in a highly competitive market.