Clemens Hammacher
PhD Student / Research Assistant

Who I Am

Since February 2011 I am working as a scientific assistant at the Software Engineering chair. Already before the start of my dissertation, I did a lot of work and research in the field of Software Engineering.

Research

My research interests lie in the field of dynamic analyses, and program transformations especially for automated parallelization of sequential programs. My belief is that in irregular programs (which is the majority of today's applications) it is very hard to prove the independence of certain parts in order to parallelize them. So with static analyses alone you are unable to extract substantial parallelism, and this is where dynamic techniques will prove beneficial. By combining for example static identification of beneficial regions with speculative execution and contention management (for example by using STM techniques) we may leverage a much larger amount of the available parallelism.

Publications

These are my most important publications:

• Thread-Level Speculation with Kernel Support
  by Clemens Hammacher, Kevin Streit, Andreas Zeller, Sebastian Hack.
  Proc. 25th International Conference on Compiler Construction (CC), March 2016.
• Sambamba: A Runtime System for Online Adaptive Parallelization
  by Kevin Streit, Clemens Hammacher, Andreas Zeller, Sebastian Hack.
  Proc. 21st International Conference on Compiler Construction (CC), March 2012.
• Profiling Java Programs for Parallelism
  by Clemens Hammacher, Kevin Streit, Sebastian Hack, Andreas Zeller.
  Proc. 2nd International Workshop on Multi-Core Software Engineering (IWMSE), May 2009.

You can find and download all my publications at the chair's publications page.

Teaching

• Speculative Parallelization (Seminar, Summer 2014)
• Compiler Construction (Core lecture, Winter 2013/2014)
• Software Design Lab (Practical Course, Summer 2011)

Projects

For my bachelor's thesis I implemented a dynamic slicer for Java, called JavaSlicer. It is available as open source, and although still unstable already used for several research projects.

Contact

The easiest and safest way to reach me is via Mail. You can also pay me a visit in room 2.16, building E9 1, or phone me at +49 (0)681 302-70983.

Impressum ● Datenschutzerklärung