Used in research on key tasks for the commercialization of ethanol, a next-generation biofuel.

NVIDIA (CEO Jensen Huang) announced that the Titan supercomputer, accelerated by Tesla GPUs, plays a key role in the next-generation biofuel development research being conducted by Dr. Jeremy Smith of Oak Ridge National Laboratory.

Ethanol, which Henry Ford, the founder of the American automobile company Ford, once called the "next-generation fuel," is a biofuel extracted from corn that can be utilized in a wide range of fields as a renewable transportation fuel. However, the biomolecular structural characteristics of ethanol act as an obstacle to mass-producing it at a level where it can generate economic value. Dr. Jeremy Smith, a scientist at Oak Ridge National Laboratory under the U.S. Department of Energy (DOE), conducted research to overcome this problem using the supercomputer Titan.

Dr. Jeremy Smith and a research team at the BioEnergy Science Center (BESC) under the U.S. Department of Energy utilized the supercomputer Titan to conduct the largest and most complex biomolecular simulations in history at a very detailed level.

The focus of this research is lignin, a major component of plant cell walls. A critical part of the process in which plant raw materials are converted from monosaccharides into the biofuel ethanol is the breakdown of cellulose. However, the plant cell walls composed of lignin play a role in blocking cellulose-degrading enzymes. Therefore, extracting lignin is key to cost-effective ethanol production.

Dr. Jeremy Smith's research team was able to conduct large-scale observations on the binding properties of lignin by simulating a biomass system composed of 23.7 million atoms using the Titan supercomputer. The team combined the insights gained from this simulation with other experimental results to develop new chemical pretreatments and improve biofuel yields.

The study results showed that in an experimental mixture of tetrahydrofuran (THF) and water and a 250,000-atom model composed of lignin, the mixture acted as a barrier between the water and the lignin. In other words, if a solvent such as THF is used as a buffer, lignin can be easily removed during the biofuel processing.

This model also revealed that lignin exhibits selective binding preferences and shows a high preference for crystalline cellulose fibers. This detailed information would have been impossible to ascertain without large-scale computer simulations conducted on Titan, a supercomputer equipped with NVIDIA Tesla GPUs.

Dr. Jeremy Smith's research team also observed the mechanism of lignin's action using a relatively small-scale simulation model of 200,000 atoms. By comparing lignin with hemicellulose, another major component of plant cell walls, in natural and genetically modified biomass systems, the team derived insights into how the hydrophobicity or water repellency of modified lignin can increase biofuel yields. These insights are expected to be useful for subsequent research aimed at making plants easier to decompose during the biofuel production process.