Professor Michela Taufer has deployed "Geronimo", a HPC Linux Cluster (triple NVIDIA Tesla S1070) from Penguin Computing making use of GPUs (Graphics Processing Units) for scientific computation.  Geronimo's 12 GPU's deploy 2880 processing cores and there are an additional 48 conventional CPU's. The floating point peak performance is 12.4 teraflops per second.  For comparison, the largest Beowulf clusters on campus have peak performance of about 2 teraflops per second.  Taufer's Global Computing Lab designs, implements, and tests high performance parallel algorithms for molecular modeling.