Keynote Raghu Belur

Thank you, Josh. Good afternoon, everyone. Thank you to Renew Wisconsin for giving me this opportunity to speak. I'm not shy about sharing my thoughts about the industry and how I see it evolving. We are in the midst of great change, and along with great change comes great opportunity. When you look at everything going on in our world today, there's a lot of noise—geopolitical, policy, 25D, 408, this and that. But as an engineer, I always think back to first principles. Yes, there's all this noise, but what is fundamental in change? I think about our energy infrastructure. Our energy infrastructure has to evolve and change. I also think about how critical our industry is. It's not just that we all want to win; it's an imperative to win, not just for our industry but for everybody's benefit. Let's start with technology. Enphase was my fifth startup, which is normal for someone who spent their whole technology career in Silicon Valley. When we started Enphase, it was the classic two guys and a dog in a basement, venture-funded. Our view was always that decentralization or distributed architectures always win in the long run—for cost, performance, and reliability reasons. If you look at all the technology existence proof out there—mainframes to PCs, plain old telephones to smartphones, data centers built using thousands of replicable blade servers—you see that more and more intelligence gets pushed deeper into the network. The endpoints are very intelligent and make autonomous decisions. Now look at our energy infrastructure. It's completely centralized: big generation, transmission, distribution, and the home. The home has not been that intelligent, and that's what our industry has changed. Our industry has started the change where the home becomes the unit of intelligence. The home makes autonomous decisions—it produces its own energy, consumes it, stores it intelligently, optimizes locally. But it's also a hyper-networked system that helps optimize the operation of the grid. That's the grand vision. We are only about 4% into that vision—the penetration of renewables in the US. We have 96% more to go. That's where we started when we founded Enphase. Why is everything centralized? All other technologies have evolved to decentralization. Here's a 125-year-old technology that hasn't evolved. It needs to evolve, and that evolution has begun. The second thing is it's not about solar anymore. We are not a solar industry; we are an energy industry. Energy means more than solar. Solar is still the hero of the show, but it's about solar, batteries, EVs, electrification of the home, managing that whole energy system, and all the software that ties it together. It's not just optimizing the home alone but participating in the operation of the entire grid. What is the alternative? There isn't one. This is the only choice. Homes are electrifying. An average US home consumes about 30 kilowatt-hours of energy. Every EV consumes 15 more kilowatt-hours per day. Two EVs double my home's consumption. Add a heat pump—that's another 15 kilowatt-hours. Heat pumps are about 4x more efficient than fossil fuel-based heating. So we have massive and growing demand on home energy. Five or eight years ago, we weren't thinking about AI data centers, especially training data centers. They consume massive amounts of energy and are very demanding. Their power demand fluctuates from 20% to 100% four times a second. If you reflect all that back to the utility, it's going to be very stressful. Our utilities are really stressed because they're getting pressured from both the data center end and from homes. In the backdrop, we have to harden our grid. We've got 100-year weather events taking place every year. All of that requires massive investments, and in a volumetric pricing environment, those investments are reflected in your utility bill. That's why we're seeing significant increases in utility bills, and that's not going to stop. We talked about energy demand in the home going up 2x or more, and that demand is very peaky. On the grid side, a gigawatt data center—data centers are specified in power, not in exaflops. They're specified as gigawatts. A gigawatt data center is nothing. People are talking about the Stargate project—five gigawatts, 15 gigawatts. Every gigawatt of data center costs $50 a watt—$50 billion for one gigawatt—and it's 640 acres. Insane amount of power needs. As a snapshot of time, it's a zero-sum game. That power has to come from somewhere. So should utilities focus on data centers or on the home? It's very messy. Now let's focus on the home. We can help. Not only can we help, we are really the only solution. When power demand in the home is going up as much as I said, how do you solve it? To support it, you'd have to rip and replace the distribution infrastructure—a hundred billion dollar proposition and extremely disruptive. You don't have a choice. It's solar and batteries in the home, EV chargers, software, heat pumps, and energy management. That is the solution. But take it a step further. Not only do we help optimize the home's energy consumption, but we can also help the grid by actively participating in energy markets—VPP or flex loads. Today, people say, 'I have a battery, so that's my VPP contribution.' But we've got to think bigger. What else can contribute as a flex load? Behind-the-meter resources can be really powerful in helping the grid. It's not just stationary batteries; it's loads, EV chargers, bidirectional EV chargers. By the end of this year, EV chargers will be fully bidirectional. That gives our industry tremendous power—pun intended. We are the only choice everybody has in all of this noise. I really want to emphasize that while there's a lot of noise and a lot happening, our industry's role is very critical and important. Sometimes we tend to underestimate that. Now, that doesn't mean we're not facing challenges. We are. But I've been in this for 20 years. I passionately dislike the word 'solar coaster.' I don't agree with it. A roller coaster leads me to think we're not in control, but we are. Let's look at some changes. 25D has expired. California with NEM 3.0, which expired in 2023, was a very disruptive event—massive demand pull-forward and then the whole market slowed down. At that time, I was probably one of the few people who went out and said, 'Hey, this is the best thing that could have happened to California.' Yes, real jobs were lost, I accept that. But it was the best thing for California and our industry in the long run. And I was right. California revenues are at the highest right now. Why? Because now we have 100% solar attached, and we're no longer beholden to anybody saying they're going to take it away. It's done. When we think about 25D and the impact, yes, it's painful. But I always say these incentives—whether an investment tax credit or NEM—need to be a catalyst and not a crutch. Other industries have been getting incentives for the last 100 years. Let them be. What we want is a level playing field. Let free market drive it. Let the best technology win, and we know our industry's technology will win. So while it feels painful that 25D is going away in a place like Wisconsin, I think in the long run it's not a bad thing. We are no longer a political football; they can't hold it over us. I dealt with it for the last 18 months—endless trips to Washington DC, and it was painful. I'd rather be in California inventing things than traveling to DC trying to convince politicians, who are very reasonable when they speak with you, but then politics takes over and all hell breaks loose. I really think the sooner the government's out of the energy business, the better it is for our industry. We should never underestimate the importance and criticality of our industry. I'm just going to take a quick peek at my notes to make sure I didn't miss anything. In some sense, I'm just going to conclude here. The changes are so dramatic. The role our industry plays is vital. It's critical. Winning is not just a nice thing; it's an imperative. We have to win for the betterment of everybody because our industry is the only solution. It's not about solar anymore; it's about energy. Everything feels right about it when you think about it on a first principles basis. Technology evolves in this direction. This is the only solution. We are the economically most viable solution, the most reliable solution, the best solution for the homeowner. And we can also help make the grid better because we are no longer simply attached to the grid; we are actually integrated deeply into the operation of the grid at multiple levels. Thank you.

Thank you. One way to think about it is simply supply and demand. Pricing signals are very powerful to shape behavior. Having said that, we should always think about this as a problem of abundance, not scarcity. We have plenty of sun. So pricing signals become really important. Historically, pricing signals have been pretty flat, so they don't really shape behavior. But now you're starting to see in places like California, you have time-of-use rates. For example, where I live in Northern California in the Bay Area, my daytime rates when I buy energy from the grid are about 25 cents a kilowatt-hour. But then summer afternoons from 4 p.m. to 9 p.m., I could be paying 60 cents a kilowatt-hour. In San Diego, from 4 p.m. to 9 p.m., people pay 82 cents a kilowatt-hour. So massive amount of energy, but it's supply and demand because at that time demand is very high and supply is constrained. California is about a 40-gigawatt grid. So pricing signals can definitely shape behavior. San Diego has the biggest battery attached. Why do they do that? Because the battery gets the chance to arbitrage the rates. In the middle of the day, I charge my battery, and between 4 p.m. and 9 p.m., I do not buy from the grid. In NEM 3, they did something really interesting. They said if you export energy in the middle of the day, the utility pays us zero to 2 cents a kilowatt-hour. Why? Because they're swimming in electrons—there's plenty of energy available. But if you buy energy between 4 p.m. and 9 p.m., you could pay 82 cents a kilowatt-hour. On the other hand, if you export energy between 5 p.m. and 7 p.m., they were paying at one time $3 a kilowatt-hour. It was worth it. It was better to pay the endpoint $3 a kilowatt-hour than start up a peaker plant. It's like pure economics. If you look at what's happening in Europe, it's really interesting. They've already moved to dynamic tariffs. At 11 a.m., our system receives a schedule that starts at midnight that says here's your hourly buy and sell rate—completely dynamic. It's exactly how the utilities buy and sell energy. It's a day-ahead market. And then our systems—in our case, I'm speaking for the Enphase system—look at the rates and adapt and figure out: when do I charge the battery? When do I discharge the battery? When do I charge your car? Soon, when do I discharge the car? When do I turn on your heat pump? Because there are times when you get negative rates. Utilities may have so much energy that they pay you to take that energy off the network. On top of that, there are requirements for imbalance. So balancing out—if they don't predict supply and demand correctly, it creates a lot of imbalance, which is again disruptive to the network. But my point being, energy becomes a really manageable market. And if you have intelligent software sitting there looking at all of those variabilities, you can deliver fantastic economics to the homeowner. Delivering fantastic economics to the homeowner means their payback reduces, ROI goes up, they buy more systems. So it's a complete win-win in that situation. That's where the markets are evolving towards—more and more either fixed signals like they have in NEM 3, but very soon you'll start seeing day-ahead pricing, real-time pricing for electrons, and like it should be. Systems have become really intelligent. AI helps a lot in making those determinations on how the system should behave to optimize the homeowner's needs.

Correct. 25D is the tax credits that were available for homeowners—the 30% investment tax credit that was available for homeowners that sunsetted on December 31st.

I mean, if I understand your question correctly, I think I'm talking about the energy infrastructure and the technology that's driving the energy infrastructure. And so I think for the most part, companies like Enphase are doing a lot in not looking at it as, 'Okay, I'm going to put more solar and solve this problem.' That's just not how you solve the problem. You've got to look at it more holistically. You've got to look at it like I'm solving the energy problem and I'm solving the network problem. That's what I mean.

It will. I think more and more there are a couple of driving factors. It's about timing. First and foremost, people have to recognize the value of that flex load. That load is a very, very powerful flex load. I call it a load, but it's actually both—it's a flex resource. Not only can I be intelligent about when do I charge, what rate do I charge at, which kind of electrons do I use, do I buy from the grid and charge my car, do I do green charging—which means I only take my excess solar and charge the car—plus, hey, what else can I do? Can I take what's charged in my car and dispatch it to the grid when needed and get paid for it? Can I use the energy that's in the car and service my home in the event of a grid outage? So if there's a grid failure because of a bad weather event, I don't have solar, I don't have stationary batteries, but I have my EV, I can then power my home with an EV. So it's a question of timing. One is recognition of the value that it creates. And second, there are some practical things like we needed a standard for bidirectional charging. So that standard's there now—there's ISO 15118, which is a communication standard, a nice handshake between the system and the car. And then finally, we needed the car guys to open up the car to enable us to discharge the battery, and that's also happening now. So it was a question of timing. The technology existed. It was take the technology that we have that we put in a stationary battery and adapt it to the car, and that's what we have done. There's a very nice video on the Enphase website if you go take a look at it. We've done a lot of testing with multiple cars and shown how nicely it works. It's really nice.