Appentra is pleased to announce the release of Parallelware Trainer 1.1, improving the provision of accessible HPC and parallel programming training using OpenMP and OpenACC.
Appentra’s goal has always been to minimize and eventually remove the parallel software development barrier, and making parallel computing easier for everyone. Parallelware Trainer 1.1 provides an enhanced interactive learning environment, including provision for a knowledge base designed around the code being developed and information on datascoping for accurate parallel programming.
New in Parallelware Trainer 1.1 version
The knowledge base provides for a tailored learning environment by helping to identify key concepts where you still need to understand a little more. It is accessible either from the Help menu or by clicking on terms appearing in the console outputs.
Data scoping report.
Available from the parallelization configuration dialog the Data Scoping Report provides information on the variables involved in the loop or region subject to analysis and parallelization. This is then used to help accurately scope variables for OpenMP and OpenACC scoping clauses.
Loops containing calls to independent functions can now be parallelized.
Support for structs.
Loops that contain structs can now be analyzed
Parallelization of code regions.
Opportunities are now shown not only for individual loops but also for parallelization regions. Regions can consist of multiple loops and are specified by the user for analysis by enclosing the region in curly braces.