My research interests include several aspects of Parallel Computing:
Resource-aware computing
Resource-aware computing is a large umbrella covering a large spectrum of resources including hardware (cpu cores, memory and caches, link bandwidth, etc.) energy and power, even human resources (e.g. coding productivity). I am particularly interested in the performance optimization of resource demanding applications on modern execution platforms, resource-aware run-time systems and allocation strategies and performance modeling.
Parallel programming
Parallel programming models, compilers and tools are topics that I find extremely interesting and challenging. Hybrid programming models for clusters of CMPs (e.g. mixed MPI+OpenMP), multicore platforms (e.g. Cilk and TBBs) and Transactional Memory have mostly attracted my attention.
High-performance computer architecture
I am interested in novel parallel architectures such as homogeneous or heterogeneous multicore systems, manycores, clusters, accelerators etc.
Scheduling resource-demanding applications in CMP and extrement-scale architectures
Resource contention is an interesting problem in modern parallel systems at all scales. We are working on contention-aware scheduling algorithms to address the problem and increase system throughput, energy efficiency and QoS.
More details can be found here.