 Let me start with a little bit of introduction to what I think is happening in this field, and this is kind of my version of the dream of computational material science from the molecular scale. We want to take what we understand at the atomic level of electrons and nuclei and atoms. We want to combine it with what we can do now on modern computers. We want to use that to understand, to optimize, and really to discover new materials, be it for nuclear reactors or electric vehicles or what have you. The message I really want to send today is that, well, sorry, first I should say this is not done in isolation. You have to do it in close collaboration with the experiment. I won't focus on that aspect right now, but that is certainly something that is always part of it. What I want to say is that this is a dream, but it's really being realized. This is really happening now, and I want to tell you a little bit about that, and that it is such a powerful tool, it really is a major transformation for what we can do in energy science and material science in general. I think there are really three fundamental drivers that have allowed us to realize this transformation today. They are the fundamental theory that underlies the behavior of atoms and electrons, the computational power we have, and then the modeling methods that connect these two. I want to tell you a little bit about each of these. First, the fundamental theory, I don't know if all of you are familiar with this, but this is Erwin Schrodinger's equation, which he wrote down in 1927. One of the profoundest statements about this actually came from Paul Dirac, who was another great quantum mechanical theorist at the time. Just two years later, what Dirac realized was that really a fundamental change had taken place in human intellectual development. This was it, this was the equation that describes how matter works on the time and length scales, at which we will probably operate for the foreseeable future. That meant that most of physics, all of chemistry, I would add, all of material science, the basic equations are done, we're there. And he realized all he had to do was solve them. So that was really the problem that was left, and I think he had a very profound realization there. So what we've been facing since then is the problem of solving that equation in order to understand materials. So what has been a huge asset in solving those equations is obviously computers. I don't need to tell you about how transformative computer technology has been and what we can do, but I will show just one figure. This is from a beautiful website called top500.org, which I definitely recommend you go play with. It is a list of data on the top 500 computers in the world as a function of time. And this is just a plot of the performance in flops, which are basically multiplications per second, and plotted as a function of year. And this is the slowest of the 500 computers. This is the fastest, and this is the sum of all of them. Many of you, probably all of you have heard of Moore's Law, which says the transistor density doubles about every 18 months to two years in a computer. That has led to doubling of speed and performance about every couple of years. The amazing thing about these computers, and this is what we really calculate on materials design, the doubling rate for this for at least 20 years has been every 1.1 years. That means just in the last 20 years alone, we have gained 100,000 times the speed of what we can compute. This is really transformative in terms of what you can do. Now, these are the 500 fastest computers, but actually you can buy computing technology today that is really on par with this. Let me give one statistic. If you want to think about what this means in terms of these floating points, if you took every person on Earth and you hired them all to do your calculation for you, and you copied that 10,000 times, that's about what you can do with one of these computers. Let me show you the kind of things that one can do. What I'm showing here is calculated voltage of a lithium ion battery. You would have, if you used a lithium metal phosphate cathode, where the metal was any one of these nine different transition metals, depending on how you think about copper. It's called transition metals. What you see here is that the voltage ranges from about 1.5 all the way up to over 5 volts predicted by the calculations. That's the blue triangles. These crosses are actual experimental measurements, and as they're just there to show you that the calculations are extremely close to the experiment. So you really can, in this case, trust the calculations to guide your thinking without having to do the experiments. And you can see how this can be very powerful for materials design. For example, everything down here on the left side is a pretty low voltage. You're not going to make a commercial lithium-ac battery out of that. It just doesn't have enough energy stored in it. These materials up here, from Nickel and Copper, have very high voltages. Actually, they're so high, that we break down the electrolytes we like to use. That tells you that these region here are really the optimal materials, which is really manganese, iron, and cobalt, and almost all of the research that's been done on these materials over the last 15 years has focused on manganese, iron, and cobalt-based alloys. The really exciting thing about this is not just that you can do this, but I could do this calculation in a day on the computers I have. It would take you orders of magnitude more time to synthesize each one of these materials and test them experimentally. So it's an enormous speed-up of what you can do, and what kind of screening you can do. Now, inspired by that, people have continued to go on to the next level and say, well, if I can do nine elements, can I do 900? Can I do 9,000? What can we do with that? And I was part of a very interesting effort that is related to designing manganese imion batteries. Let me just quickly introduce that. Manganese imion battery is just exactly a lithium-ion battery, except you've replaced your lithiums with magnesium. Why do that? What magnesium carries two electrons? And that means, as a first approximation, you might be able to get twice as much capacity than you could get out of a lithium-ion battery. And I had the good fortune of my sabbatical working as the vice president of research at Pelion Technologies, and this is a startup company which is dedicated to using these types of computational tools to design this next-generation magnesium battery. And here's the type of calculation that we were involved in there. So this is very much the same calculation I just showed you, but on steroids. So we've done here not nine different compounds, but 9,000, 10,000, 12,000. I forgot the exact number. And what we've done is calculated what capacity and what voltage you would get if you were to make a magnesium battery out of essentially every possible material and then a lot that weren't yet discovered, but we could think up in our head and then explore computational. And if you look at this, it's actually a product of these two numbers that gives you the energy density. This is what advanced lithium-ion energy densities are like today. This entire shaded region and all the dots in it represent compounds that could potentially give you energy densities that help perform even the best lithium-ion batteries today. And what Pelion has been doing is picking out the most promising members of this space that you've been able to identify computationally and is synthesizing and testing those in magnesium batteries. Okay, with that I will wrap up. The broad message I want to say is that I think these molecular-scale computations really are transforming what we can do in energy science. And I think a place like the Wisconsin Energy Institute where we can integrate that computation with the experiment and with all the people in one place for the kind of communication is really a critical step forward for us and I think it will help enable Wisconsin to really continue to lead in these types of transformations. So with that, thank you very much for your attention.