The HPCA group pursues the optimization of numerical algorithms for general purpose processors (superscalar and VLIW) as well as specific hardware (GPUs and FPGAs), and their parallelization on both message-passing parallel systems (mainly clusters) and shared-memory multiprocessors (SMPs, CC-NUMA multiprocessors, and multicore processors). The group is involved in the application of high-performance parallel computing techniques to the solution of problems arising in control theory, computational chemistry, electromagnetics, aeronautic engineering, and scientific and engineering applications in general.
Current interests of the group also include virtualization of hardware accelerators, power-aware computing, hardware-software codesign, reconfigurable architectures, and high-speed networks and QoS.
Contact: José I. Aliaga Estellés, group leader (aliaga@icc.uji.es).