2019
September
August
June
May
March
February
January


2018
November
October
September
August
July
June
May
April
March
February
January


2017
December
November
October
September
August
July
June
May
April
March
February


Categories

All Episodes
Archives
Categories
Now displaying: Page 1
Dec 27, 2017

Most of us think of casinos or James Bond when we hear about Monte Carlo. But to today’s guest, Monte Carlo makes him think about algorithms. Monte Carlo simulations use random sampling to produce a distribution of results from which we can draw conclusions. These computational science techniques help us answer some of the world’s toughest challenges at the atomic level.

Obviously, computational Science is an exponentially-growing multidisciplinary field that uses advanced computing capabilities to understand and solve big, complex problems. It is an area of science which spans many disciplines, but at its core it involves the development of mathematical models and simulations to understand natural systems. Think about the way that we predict the weather. In order to predict the weather, scientists run a simulation many times over, randomly choosing atmospheric data and then looking at common themes across those simulations to generate an idea of what weather is most likely for a given area. When we hear meteorologists say, “We have a 60% chance of rain today", they are really saying, "60% of our simulations predict rain today.” If we elevate our thinking from mundane rain clouds to simulations of material properties, things get interesting.

According to our guest, Chris Wilmer, thanks to computational science techniques, the beginning of the 21st century has seen an explosion in the design of porous materials for a wide range of applications, from gas storage and chemical separations, to sensing and light harvesting. In this episode, Dr. Chris Wilmer describes how he designs atomically engineered materials through the application of modern computing infrastructure, thereby developing material discovery algorithms. Using this platform, he creates millions of hypothetical structures, stores these structures in databases, and then uses high-performance computing to rapidly simulate their properties. He can then validate the performance of materials through an automated workflow, providing a powerful prediction-meets-data feedback loop.