My research activity is focused on the development of SW-HW codesign strategies as well as algorithms for parallel programming support for energy-efficient large-scale parallel computing systems, such as HPC machines and supercomputers. It also includes the study and the implementation of novel mechanisms to propagate performance requirements from the application to the operating system to support dynamic parallelization and DVFS control policies for energy-reduction under performance and power constraints.
Daniele Cesarini
Expertise areas
Topics: Computer architecture, GPUs, High-performance computing, Multicore / Manycore, Networking / Distributed computing, Parallel computing, Performance portability, Resource management / Scheduling, Runtime performance, Storage / IO