Software Development
Compiler Performance Enhancements
New compiler technology optimizes execution performance directly from the Mathematica language using two key innovations: automatic code generation and linking, and automated multicore parallelism. This enhances Mathematica's rapid application development workflow to enable final deployments requiring the ultimate high-performance execution.
- Automatic generation and linking of C code from compiled functions. »
- Multicore parallelism for compiled functions. »
- Listable operation for compiled functions to thread over input arguments. »
- New options to control compilation of Mathematica code. »
- New options to control execution of compiled functions. »
- Extensive tutorials on using the Mathematica compiler. »
- Tools to study the Mathematica compiled function bytecode. »
- Compiled functions can be exported to C code for standalone operation. »
- Compiled functions running in parallel automatically synchronize access to shared resources.
- Compiled functions running in parallel use an efficient parallel memory allocator.
- Compiled functions can insert inline code from other compiled functions. »
- Compiled functions can make efficient calls to other compiled functions. »