Japan's Solar Frontier is the only true volume producer of copper-indium-gallium-diselenide (CIGS) thin-film solar panels. The company has shipped more than 3 gigawatts' worth of modules since its founding.

No other CIGS company comes close.

GTM interviewed the company's CEO, Atsuhiko Hirano, at this year's Solar Power International conference in Anaheim, Calif.

Unlike vertically integrated First Solar, Solar Frontier sticks to manufacturing solar modules only and works with partners to "form a good, controlled value chain," according to the CEO.

Earlier this year, Solar Frontier acquired the 280-megawatt U.S. project pipeline of Gestamp Solar, a developer and operator of utility-scale solar plants -- along with Gestamp's U.S. development team. Gestamp Solar was backed by the $13 billion industrial holding group, Corporacion Gestamp. Charles Pimentel, the COO and SVP of Solar Frontier Americas, emphasized that Solar Frontier remains "strictly a module manufacturer" that is making a foray into project development. "The objective is not to own the projects long-term."

The move to non-Japanese markets is a big shift for the thin-film solar vendor. Over 90 percent of Solar Frontier's 2014 production (close to 1 gigawatt) went to domestic Japanese installations, according to GTM Research solar analyst Jade Jones. But that imbalance is about to change as the Gestamp projects begin to draw the modules to the U.S. Hirano anticipates a 50-50 shipment mix between Japan and the rest of the world in 2016.

Acquiring a turnkey project pipeline gave Solar Frontier an efficient way of entering the U.S. market and joining the race to originate and develop projects before the expiry of the 30 percent federal ITC. More importantly, it establishes Solar Frontier as an active, bankable player in this market.

CEO Hirano added that the company is looking "to approach the market in a modest way with a small set of partners and a handful of strategic developers where we can bring our resources to bear." Pimentel noted that it took a "sophisticated EPC" to understand and price-in the temperature and shading benefits of thin film.

Solar Frontier has a number of projects with 2016 CODs throughout California, ranging from the 15-megawatt Lost Hills project to an 80-megawatt project development effort. Pimentel pointed out that the projects are "almost exclusively on single-axis trackers."

The CEO told GTM that module efficiency for its 170-watt module is 13.8 percent, with champion modules at 14.6 percent. The CEO suggested that the new factory's efficiency starting point would be 14.6 percent. Thin film leader First Solar builds cadmium-telluride-based solar modules and had a full fleet average efficiency of 15.4 percent and best line efficiencies of 16.3 percent in Q2 2015.

But Solar Frontier isn't competing with First Solar -- it's competing with polysilicon-based modules, and it still has a few more percentage points to go in efficiency to be able to rival that material.

Markus Beck of Siva Power has suggested that the sputtered "Solar Frontier, Stion/TSMC, Avancis, Bosch, Hulk, Samsung approaches are all rooted in the late 1980s/early 1990s Arco/Shell/Siemens approach" and have not seen a lot of innovation on the equipment side.

Solar Frontier's annual capacity is about 900 megawatts at its Kunitomi plant, with another 150 megawatts ramping up this year at the Tohoku plant. The CEO said, "Once we are up and running, the Tohoku plant will be the mother plant." He said the Tohoku plant is a technological leap with a much shorter process time at two-thirds the capex and opex per megawatt. The CEO added that module cost would approach 40 cents per watt including depreciation, and 30 cents per watt without it. He did not provide a timeline.

The CEO also confirmed reports that talks are still ongoing for Solar Frontier to build a U.S. factory at the RiverBend site in Buffalo, New York, the same location as SolarCity's highly incentivized crystalline silicon factory.