Research Group on Compiler Technology and Parallel Computing at PELAB
Contact person: Prof. Christoph Kessler
- Design and implementation of programming languages and systems
- Compiler and runtime techniques for resource-aware execution on parallel and embedded systems
- Program optimization and code generation for parallel, distributed and heterogeneous systems
- Software composition
- Dr. Andrzej Bednarski, PhD 2006
- Dr. Mattias Eriksson, PhD 2011
- Dr. Usman Dastgeer, PhD 2014
- Dr. Nicolas Melot, PhD 2017
Other former members
- Johan Enmyren, project assistant (summer 2010)
- Mudassar Majeed, external PhD student
- Amin Shafiee-Sarvestani, project assistant (summer 2012)
- Oskar Sjöström, project assistant (spring 2015)
Some current and recent research projects / -activities:
- EXCESS: Execution Models for Energy-Efficient Computing Systems
- SkePU: Auto-tunable multi-backend skeleton programming framework for multi-GPU systems
- MeterPU: Portable, generic (time, energy, ...) measurement abstraction API for GPU-based systems
- XPDL: Extensible platform description language for heterogeneous computer architecture and system software properties supporting retargetable optimizations and toolchains
- PEPPHER Composition Tool: Code generation framework for synthesizing applications for heterogeneous multicore systems from multi-variant components with auto-tuned variant selection, developed for the EU FP7 PEPPHER project
- PRT: Generic pattern recognition tool for automatic parallelization
- SeRC-OpCoReS: Optimized composition and run-time support for e-Science
- Optimized on-chip pipelining and mapping for Cell/B.E. and Intel SCC
Some former research projects:
- OPTIMIST: Retargetable integrated code generation for instruction-level parallel and embedded systems.
- RISE-PARACOMP / RISE-II-ARECOMP: Interactive invasive parallelization, with support of automatic roundtrip engineering
- Clustering and scheduling of fine-grained task graphs generated from Modelica programs
- Vinnova GridModelica / SafeModSim: WP on parallelization
- EU FP5 JOSES - Java compiler for embedded systems, WPs on DSP code generation and garbage collection for hard real-time systems
- NestStep Design and implementation of a global address space language for BSP computing
- Language, library and compiler support for PRAM computing (Fork, REPLICA)
- SPARAMAT: Pattern-driven automatic parallelization of sparse matrix computations
- PARAMAT: Pattern-driven automatic parallelization
Courses:We teach about a dozen undergraduate-level and graduate-level courses per year, in the areas of compilers, code generation and optimization, parallel programming, multicore computing, concurrent programming and operating systems, component software and software composition, and software engineering.
For a partial list, see here.
Master degree projects:For internal 30hp master degree projects (exjobb) in compiler technology, see Christoph Kessler's list of open master thesis projects.
Current and finished master thesis projects in our group can be found on this page.
Group meetings:on demand, announced via email
Some links:see the link collection
- C. Kessler's BibTEX files: see /home/chrke/tex/*.bib
papers with a (red) C code (e.g. C453) can usually be found in the boxes standing on the middle bottom shelf in C. Kesslers office.
Page responsible: Webmaster
Last updated: 2017-12-21