For evaluation purpose a benchmark of the Cholesky decomposition is needed. Implement a Cholesky Benchmark using the Flow Graph API. The following points have to be done: - [ ] Understand the Cholesky decomposition (https://de.wikipedia.org/wiki/Cholesky-Zerlegung) - [ ] Take a look into the Cholesky Benchmark of intel TBB (https://github.com/01org/tbb/blob/tbb_2019/examples/graph/cholesky/cholesky.cpp). There are 4 different Versions implemented and 2 of them using Flow Graph. The join graph based implementation is the important one for us. Try to understand it! - [ ] Answer the following questions: * What kind of Flow Graph Nodes are used and why? * Do we have suitable Node in ProThOS to express the program and which do we need to add? * Can we reach good predictability (unrollability) of the Flow Graph? - [ ] Create an issue for each missing Node Types - [ ] Implement Cholesky Benchmark
For evaluation purpose a benchmark of the Cholesky decomposition is needed.
Implement a Cholesky Benchmark using the Flow Graph API.
The following points have to be done: