Skip to Content

EmTech: Saving Solar Startups

Bill Joy believes he knows what’s wrong with solar innovation, and how to fix it.
October 24, 2011

Bill Joy helped develop the climate change and sustainability investment strategy at the venture capital firm Kleiner Perkins Caulfield & Byers, which is known mainly for its investments in companies such as Amazon.com and Google.

Solar booster: “We’re going to have to provide capital to be competitive,” says Bill Joy. “The Europeans provided it indirectly with feed-in tariffs. The Chinese provide it more directly. By comparison, what we’re doing in the U.S. is very little.”

Joy is also famous for building BSD Unix, a computer operating system that has had a big impact on the computer industry and the development of the Internet. As a cofounder and chief scientist at Sun Microsystems, he developed key technologies such as the programming language Java.

As an investor, Joy has backed several high-risk solar power companies. Technology Review talked to him about lessons learned from failed investments in solar companies, and what it will take for innovative solar companies to succeed.

TR: With the failure of Solyndra, and hard times for many other solar companies, solar investments have come under scrutiny. What’s gone wrong?

Joy: Many have been badly managed, or badly conceived. The trouble comes if you’re not good enough to make a difference. I think a lot of the solar ventures haven’t had enough differentiation. They haven’t been enough better than the trajectory of the incumbents.

[At Kleiner Perkins Caulfield & Byers] we’ve had some failures, but that’s because we’ve been aiming well ahead of where the conventional technology was going to go. The alternative is that you can try to do something that isn’t as far ahead. So then what’s the point? You’re just doing almost the same thing using immature manufacturing that everyone else is doing using mature manufacturing. That’s no recipe to win.

You mentioned failures at KPCB—what went wrong there?

Solasta was an attempt to build a three-dimensional nanostructure [for solar panels] that would’ve had better physics, and the difficulty there, I think, was fabrication. We were unable to come up with a technique for making the devices that we wanted. We didn’t have the time to wait for the new technology to make it work.

Intel’s new transistors—they’re turned up in the third dimension—I heard they spent something like $7 billion and had a thousand PhDs. It was an extremely large program, and went on for a long time. We can build patterns that are planar, but to build structures that are vertical, and that fine-scale, is very difficult. Solyndra was trying to deposit on a nonplanar surface. Doing stuff on a planar surface is easy. Anything beyond that is difficult.

You recently invested in the solar startup Alta Devices. What lessons did you apply from Solasta in considering that company?

Alta Devices is building a more traditional planar structure. Alta Devices starts with what’s essentially the perfect solar material—gallium arsenide. It has an absorption spectrum that’s almost ideal for the sun. It absorbs light much more strongly than silicon. It can be something like 100 times thinner than the silicon, so you need something like 100th of the material.

The Alta cells are lighter and thinner. They take in the light at a low sun angle better. They have better performance at temperature and better low-light performance. They have higher efficiencies so the solar cells can be smaller. Alta Devices has the world record efficiency for solar, yet measuring efficiency alone understates the advantage of gallium arsenide. The world record efficiency measurement doesn’t count temperature, doesn’t count low light. If you integrate over a day, and over temperature cycle, and over a year, you get much more energy than silicon.

What are the challenges with Alta’s technology?

It’s harder to put the manufacturing together than silicon. You can’t just take silicon ingots that were meant for making semiconductors and slice them and stick them on a piece of glass. Making thin-film gallium arsenide cells is more like [the manufacturing done at] Intel. It’s much more like building a perfect transistor: conceptually you’re building one giant one.

Will the manufacturing be difficult to scale up?

With solar, the individual machines and the individual technology of manufacturing is difficult, but scaling up is not as difficult. It’s a different kind of scaling up. You get to a size for a piece of manufacturing equipment which is large enough, and in a single factory you might have 16 of the machines, but you don’t need a machine that’s 16 times as big.

Will you need to find niche applications to bring the technology to market, while you bring costs down?

You might find a specialty market where there is a higher price. If you’re very light and very efficient, that’s a great thing to put on a car. You need to be flexible, you may need to be curved in two directions, and there’s only so much area, and you don’t want to be heavy. So a high-efficiency thin-film is an example for there.

There are lots of applications. If you want to put stuff on a roof, you want to reduce the weight. That’s one of the things that Solyndra was trying to do. Cut the dead load on the roof—with these tube-shaped solar cells. You can also do it with a very lightweight and thin material, and that’s probably a more direct and a simpler way to do it.

But the fact is that the new technology can be both cheaper and have new attributes.

Are there any government policies that are essential to scaling up the technology?

The policies help. But we’d better be economically viable with limited policy support because unfortunately, in the United States, the policy is really uncertain.

In this country we are very conflicted about whether the government should try to help companies get to scale. But if we want to have the jobs in this country, we should support this. Are we glad we have the PC industry? Yes. Did the government support it? Yes. By doing DARPA, and doing the research.

In the energy space we’re going to have to provide capital to be competitive. The Europeans provided it indirectly with feed-in tariffs. The Chinese provide it more directly. By comparison, what we’re doing in the U.S. is very little.

As the politicians use the Solyndra debacle as a political football, it’s not helping the country. We need to find ways for these companies to get financed. One wasn’t well conceived, and it’s the nature of these things that some of them will fail. But maybe we shouldn’t have ever had the expectation that we couldn’t lose our money.

Maybe the government should ask for some equity as well. They could loan the money and get some stock. And with the ones that succeed, the stock will pay off the ones that don’t. That’s kind of what happened with the auto bailout, right? They got a bunch of stock in GM, made a lot of money off of it. I should say “we” made a lot of money.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.