Dynamic taint propagation in hardware accelerators (in collaboration with NYU);
Security-aware design of cyber-physical systems (in the context of the CERBERO H2020 project);
High-level synthesis of high-performance accelerators
Development of BAMBU, an open-source framework for research in high-level synthesis;
Accelerator memory microarchitecture to support the high-level synthesis of unmodified C code with complex data structures and memory operations;
Comparison with state-of-the-art HLS tools (in collaboration with TU Delft and University of Toronto);
Generation of dedicated memory subsystems based on polymorphic register files for the Maxeler platforms (in collaboration with Maxeler and Chalmers University);
System-level methodology for the design and optimization of the accelerator local memories in high-throughput accelerators (Mnemosyne, open-source CAD tool);
Hardware/software co-design with limited hardware resources
Co-design methodologies based on Ant Colony Optimization for hardware/software partitioning, optimizing tasks and communications with several constraints on the application and the architecture (e.g., limited resources) (in the context of the hArtes FP6 project);
Extensions to BAMBU for multi-objective design space exploration based on genetic algorithms;
Path-profiling techniques and estimation methods for supporting automatic parallelization;
Virtual prototyping of heterogeneous architectures;
Self-adaptive and reconfigurable systems
ACO-based methodologies to determine the characteristics of the reconfigurable regions (in the context of the FASTER FP7 project);
High-level analysis and design of heterogenous systems taking into account the effects of partial dynamic reconfiguration;
FPGA prototyping of SoC architectures with accelerators processing massive data sets and featuring dynamic voltage and frequency scaling;