In the recent decades tremendous progress was achieved in many fields of scientific computing such as machine learning, genome sequencing, numerical simulations, to name but a few. This amazing success was not only made possible by the continuous development of the algorithms, but also by the constant improvement of powerful programming languages and development techniques to manage the complex algorithms. However, many scientific software projects are still hard to maintain or failing because of deprecated programming languages, libraries, or features or disregarded well-established development techniques.
The purpose of BrainTwister is in the development and provision of high performance software in the field of scientific computing, while at the same time special emphasis will be placed in the usage of state-of-the art programming language features and development techniques without compromises.
This includes for example:
- Version control via Git
- Project management via GitLab and GitHub
- Continuous integration via Jenkins
- Development environment via Docker
- Extensive software testing
- Software metric control
Thematically, the main focus is at the moment in the development of BlasBooster, a linear algebra library to combine dense and sparse algorithms.
I hope that my short outlook has made your interest grow to find out more and I am looking forward to received your feedback.