Researchers at the University of Illinois at Urbana-Champaign (UIUC) have made a major breakthrough in the battle against HIV. They’ve managed to make an atom by atom simulation of an HIV Virus and have thus managed to completely map out the structure of the HIV capsid, a protein shell that protects the virus’ DNA and is key to it’s virulence. The research team believe that while no current HIV drugs target the capsid shell, it is an attractive target for future treatments: It has been discovered that capsid disruption as caused by a protein in Rhesus monkeys, has led those monkeys to become immune to the virus. The team made their breakthrough with the help of Nvidia Tesla GPUs sitting in the Blue Waters super computer.
Blue Waters is a Cray XK7 super computer and houses 3,000 Nvidia Tesla K20X GPUs. The HIV study marked the largest simulation ever published for Blue Waters, mapping out the behavior of over 64 million atoms in the HIV virus. “It would have been very difficult to run a simulation of this size without the power of GPU-accelerated supercomputing in the Blue Waters system. We started using GPU accelerators more than five years ago, and GPUs have fundamentally accelerated the pace of our research,” said Klaus Schulten, professor of Physics at the University of Illinois.
Sumit Gupta, general manager at Nvidia’s Tesla Accelerated Computing Business Unit, explains that GPU accelerated super computers are able to run more complex tasks than older CPU-only powered computers did. “GPUs help researchers push the envelope of scientific discovery, enabling them to solve bigger problems and gain insight into larger and more complex systems. Blue Waters and the Titan supercomputer, the world’s No. 1 open science supercomputer at Oak Ridge National Labs, are just two of many GPU-equipped systems that are enabling the next wave of real-world scientific discovery,” he explains.
Nvidia developed the CUDA programming language several years ago with the purpose of helping super-computer like Cray make use of both GPUs and CPUs in their systems. Today, the majority of super computers make use of this kind of parallel processing to speed up computations.