SolarCity stock is up 500 percent from its IPO price. There's asolarrecovery happening that is not only lifting the industry’s dealers and installers -- the stocks of the leading panel manufacturers such as First Solar, SunPower and Yingli are also up 300 percent on the year.

Deutsche Bank reported that by the end of 2013, two-thirds of the world's markets could economically use solar without government subsidies. And this growth will likely continue for decades.

So, the good times are back? 

Not so fast.

There is a dilemma in the numbers.

The industry’s growth has been based on continually declining prices for solar panels. Low prices are good for the installers, but bad for panel manufacturers. In fact, despite rising share prices, almost all of the module makers are still losing money. In China, the top solar manufacturers have a combined $16 billion in debt. Their bankers are beginning to believe they might never get paid interest and might even have to forego the principal.

So if growth is to continue, who is going to loan these indebted companies the money to build new factories? The panel makers need higher prices to return to profitability and pay back their debt, but the market needs lower prices to sustain its growth. What will give?

Having seen this before in the semiconductor industry, I believe the probable scenario is that we will split the difference. Prices will muddle along, moving slowly downward to maintain growth, and what little margin is eked out will be 100 percent consumed by debt payments. High growth with no profit means we could be entering an era of “profitless prosperity” for the solar industry.

A solution exists, however. The current solar manufacturing capacity was built years ago, but designed even further back in the era of solar hobbyists. Call this Solar 1.0, the megawatt era of solar.

With demand set to reach between 45 and 55 gigawatts next year, we are truly entering the era of Solar 2.0.

Now we just need factories designed for gigawatt capacity.

If this is done properly, manufacturing costs can decline fast enough to get margins back to the 20 percent to 40 percent range seen in the past. But it will require innovation -- not just in R&D, but also in factory design.

The problem is that the current solar factories expanded so fast that their owners did not really scale operations. To see how margins can increase through economies of scale, we need only look as far as other semiconductor-related industries such as flat-panel displays and integrated circuits. (By the way, solar is a semiconductor-related industry. A solar cell is merely a diode, which is a semiconductor device.)

Each of these industries has scaled by increasing substrate size. Flat-panel displays enlarged substrate size by over 50X, and integrated circuits by over 20X. What has solar done? Would you believe less than 2X? If these other industries scaled the way solar has scaled, there would be no iPhone 5, or any iPhone at all.

The solution is simple: scale, don't replicate.

Don’t just build more small, unprofitable production lines, but scale them up from their current 25-megawatt to 30-megawatt level to a 250-megawatt to 300-megawatt size. This is possible, but it takes some innovation in factory, equipment and process design.

Fortunately, innovation is an American strength. In fact, the U.S. is home to most of the world’s leading semiconductor equipment manufacturers. Maybe the key to the gigawatt-scale solar factories of the future resides not in China, but right here in Silicon Valley.

The reason scaling works is that these high-speed production lines would have 10X the capacity, 10X the output, but only 2X to 3X the cost. This math has been proven over and over again in flat-panel display and semiconductor factories. Not only is factory cost reduced, making financing easier, but production costs come down as well. The gigawatt-scale factory of the future is both the solution to profitless prosperity and an opportunity to bring solar manufacturing back to the U.S.

One huge challenge remains: Who will finance the development of these high-speed, gigawatt-scale factories?

The good news is that much of the research work has already been done by the flat-panel display industry.

Still, investment is needed, and many U.S. investors are sitting on the sidelines.

Unfortunately for the U.S., huge investments are being made in China. Over $100 billion has been allocated to solar development in Saudi Arabia. So investors will be found overseas, but at a heavy cost to America’s competitive infrastructure.

Can we really afford for other countries to control the energy source of the future?

***

Brad Mattson is CEO of thin-film solar company Siva Power.