[HiPEAC-announce] release of cTools to automate program/compiler/architecture optimization

[Grigori Fursin]

Dear all,

Finally, we managed to release/update all collaborative R&D tools 
at http://ctuning.org to enable collective optimization of computing systems:

* Fully redesigned and documented Interactive Compilation Interface (ICI) v2.0 for GCC 4.4.0 
synchronized with the official plugin GCC branch. It is used to transform GCC
into research compiler using external plugins. Several branches have been created
for further development during this summer, i.e. providing function cloning support
to enable run-time adaptation, providing fine-grain selection of transformations
(including polyhedral ones using GRAPHITE) based on machine learning, providing program 
instrumentation support. More information/downloads: http://ctuning.org/ici

* MILEPOST GCC 4.4.0 pre-release version - it's a combination of ICI and static feature
extractor to automate program optimization and compiler tuning using machine learning.
Follow download links at http://ctuning.org/ici	

* Continuous Collective Compilation Framework v2.0 - it's a plugin-enabled framework
to automate program and architecture optimizations (i.e. iterative search for good 
program and architecture optimizations in a feedback-directed manner) 
and gather various static and dynamic optimization profile data in a Collective Optimization
Database.
More information/downloads: http://ctuning.org/ccc

* Collective Benchmark(MiDataSets) v1.0 - it's a collection of open-source programs 
with multiple datasets assembled by the community to enable realistic benchmarking 
and research on program and architecture optimization. 
More information/downloads: http://ctuning.org/cbench

We also updated Collective Optimization Database with various optimization cases for Intel and AMD 
processors and started collecting data to compare different compilers including GCC, LLVM, Open64, 
Intel, etc. COD is used to enable sharing and reuse of optimization knowledge. It is also used
to improved the quality and replicability of research.

Online optimization predictor based on static program features (from MILEPOST GCC) and optimization
cases from COD is also available online at http://ctuning.org/cdatabase

We would like to thank cTuning community for feedback, help and support! 
You are welcome to join this effort to automate program, compiler and architecture design 
and optimization at http://ctuning.org/community

Yours,
Grigori Fursin

========================================
Dr. Grigori Fursin, INRIA, France
http://fursin.net/research