Sites Inria

Version française

COMPSYS Research team

Compilation and embedded computing systems

  • Leader : Alain Darte
  • Research center(s) : CRI Grenoble - Rhône-Alpes
  • Field : Algorithmics, Programming, Software and Architecture
  • Theme : Architecture, Languages and Compilation
  • Partner(s) : Université Claude Bernard (Lyon 1),Ecole normale supérieure de Lyon,CNRS
  • Collaborator(s) : U. LYON 1 (UCBL), ENS LYON, CNRS, INRIA

Team presentation

Compsys is a common research project-team of Laboratoire de l'Informatique du Parallélisme (LIP, Lyon) and Inria. The team is located at Ecole normale supérieure de Lyon (ENS Lyon).

The goal of Compsys is to develop compilation techniques, more precisely code analysis and optimization techniques, for programming or designing embedded computing systems. Compsys focuses on both low-level (back-end) analysis and optimizations for embedded processors and high-level (front-end, mainly source-to-source and polyhedral) analysis and transformations for hardware (programmable or reconfigurable) accelerators. The main characteristic of Compsys is its focus on combinatorial optimization problems (graph algorithms, linear programming, polyhedra) coming from code optimization problems (register allocation, memory optimization, scheduling, automatic generation of interfaces, etc.) and the validation of these techniques in the development of compilation tools.

Research themes

  • Program analysis and transformations (mainly at source level) for high-level synthesis of hardware accelerators and for compilation to programmable accelerators. Specification of communicating processes, interface generation, memory optimizations, communication optimizations, code generation for FPGA.
  • Code generation and optimizations (mainly back-end optimizations) for embedded processors. Aggressive and just-in-time optimizations (e.g., register allocation), intermediate representations (e.g., static single assignment form), program analysis.
  • Development of optimization tools (mainly polyhedral optimizations) for scheduling, code analysis, memory optimizations, simplifications of affine expressions, code generation.

Keywords: Compilation Embedded computing systems Code transformations & optimizations Aggressive and just-in-time compilation High-level synthesis Polyhedral techniques