Homepage

Welcome to the Parallel Computer Architecture Group Home Page. All the people in this group are working in the Computer Engineering Department (DITEC) of the University of Murcia.

logo-grupo.jpg

The Parallel Computer Architecture Research Group (GACOP) of the University of Murcia began in 1995, being now fully consolidated and being one of the main pillars of the Faculty of Computer Science. Currently, the group consists of 4 PhD members and around 7 PhD students and a varying number of master and visiting students.

The research of our group has been always centered around the topic of Supercomputing, with greater emphasis in the following lines:

* 1990 - until now: Interconnection networks: Routing, switching, hot-spots, and fault tolerance in high performance and throughtput interconnects.

* 2000 - until now: Computer architecture: Multiprocessor on chip (CMPs): coherence protocols, energy consumption, fault tolerance and virtualization.

* 2008 - until now: Code modernization: For CMPs and GPUs architectures, specifically in the acceleration of bio-inspired algorithms for application in the field of Bioinformatics.

* 2015 - until now: High Performance Computing for Deep Neural Networks: Training, inference, precision, memory costs.

The philosophy of the group is to contribute high-quality scientific research and be in permanent contact with national groups whose topics are similar to ours, as well as with some of the best-known European and American groups. As a result of this strategy, a significant percentage of the produced publications have emerged through collaborations with other groups. These works have been published in prestigious international journals (amongst IEEE-edited are TC and TPDS, and amongst Elsevier editions are JPDC, Parallel Computing, FGCS or JSA) and high-impact international conferences (HPCA, IPDPS, HiPC, ICPP, SC, ICS, EuroPAR, etc).