Less is more.
The aiPop optimizer suite was developed to reduce the code size, improve the code quality, and, optionally, speed up the execution of assembly or object files produced by a C compiler.
Control flow graph of a C16x/ST10 application
before and after optimization with aiPop
Featured optimizations
- Functional abstraction (reverse inlining) for common basic blocks
- Tail merging for procedures
- Interprocedural constant propagation
- EXTP optimizations
- Dead code elimination
- Optimizations based on data dependence analysis
- Peephole optimizations
- Loop invariant code motion on the assembly level
- FAR data access optimizations
- Superblock optimizations
- And more
Unparalleled compaction rates
| Benchmark reference customer application | Saved |
|---|---|
| Small (64kB) airbag application featuring highly hand-optimized code | 5.30% |
| Large (1MB) mobile phone application | 20.39% |
Further features
- Automatic code size reduction based on abstract interpretation and advanced pattern matching techniques.
- Reduction of the execution time (optional).
- High performance even for global optimizations thanks to a sophisticated pattern database.
- A graphical user interface puts all optimizations at your fingertips.
- aiPop marks all changes in the optimized assembly files.
- aiPop enables a trade-off speed vs size.
Most recently added features
- Customizable grouping of command-line parameters.
- Even higher compaction rates thanks to superblock optimizations.
Why do you need aiPop?
The
size of compiled C code is becoming increasingly critical in embedded
systems, where the economic incentives to reduce ROM sizes are very
compelling.
Tests of aiPop on complete reference customer applications showed overall code size reductions of more than 20%. Compacting code by 20% allows 25% more functionality to be packed into a flash memory of the same size.
Reduction of code size directly translates into reduced memory requirements and reduced hardware costs.
As opposed to file compression (zipping), aiPop’s post-pass optimizations do not change the functionality of your system.
aiPop supports incremental program verification, which is especially important for safety-critical applications.
Tedious and error-prone hand-crafted optimization is a thing of the past. With aiPop, you can easily select from and combine dozens of optimization techniques, trade size for speed and vice versa. Thus, saving considerable effort and expense.
aiPop can be easily integrated into established tool-chains. It was designed and implemented after conducting in-depth studies of customer requirements and in close dialog with major developers of embedded applications such as Siemens ICM. Since 2000, aiPop-compacted software runs in millions of mobile phones worldwide.
Availability
aiPop for C16x/ST10 optimizes assembly files in .src format as produced by the TASKING C compiler for C16x/ST10 microcontrollers. It is available for Windows and Linux.
Optimization of hand-written code is supported to some extent provided that it is similar to compiler-generated code.
Versions for several other target processors and compilers have been temporarily discontinued or are currently under development. Do not hesitate to contact us for more information.
In the press
The article “Post Pass Code Compaction at the Assembly Level for C16x” (PDF, 79kB) was published in CONTACT, Infineon Technologies Development Tool Partners magazine, vol. 3, issue 9.
The
article “Liebling, ich habe den Code geschrumpft”
was published in the magazine DESIGN&ELEKTRONIK, issue 1/2001.
The article “Schrumpfkur für
Computer-Chips” was published in the Saarbrücker Zeitung, issue 106/2003.
Trial version
If you would like to receive an
evaluation version of the aiPop code compaction suite, please print
out, fill out and sign the evaluation form (PDF, 75kB) and send or fax it to AbsInt.
If you have any questions, let us know.