The opportunity to undertake the modernization of development processes can be brought about by
- the desire to increase your productivity, reduce your maintenance, focus on your assets, capitalize on your developments
- requirements for new formulations and sensitivity calculations
- requirements for very fast assembly.
This is no doubt the opportunity to consider the use of NAVPACTOS, a language specially developed by Naupacte for these purposes. In fact, by removing the codes from their computational part, NAVPACTOS considerably lightens their development. What about performance? Far from being degraded, assembly speeds in particular are unprecedented.
The programming of an assembly by Finite Elements is an example where NAVPACTOS stands out for its simplicity and performance.
Simplifying rewriting is not the only advantage of this language, as it opens the way to generic programming with high reusability. Indeed, its power comes from being able to transfer formulations – or calculation definitions – thanks to its formal tensor calculation engine. Consider the features that handle such formulations: assembly, fields, meshes, algorithms like Newton’s, and therefore the higher-level modules that use them. All code based on NAVPACTOS benefits from the concept of formal calculation.
One possible opening allowed by rewriting with NAVPACTOS is the development of a parametric optimization algorithm using gradients automatically.
NAVPACTOS is therefore a tool to consider when the question arises of modernizing its calculation code.
With NAVPACTOS, you can easily compose the formulation and the corresponding system construction from the specifications of a numerical model, read for example from a command file, and then calculate and update them with a simple call. A dedicated language written in simple and expressive C++ allows you to follow the mathematical model.
Simply compile as usual your NAVPACTOS construction and resolution program. The outstanding performance of the calculations requested is delivered automatically thanks to a new method based on a few thousand internal rules.