Nowadays, Graph data are very important and omnipresent. Due to the fast growth of Internet, very large data are being more and more produced. In such a case, a cluster of machines is commonly used to store data in a distributed manner, and to process them in parallel. Many platforms and solutions exist that can process these data transparently. However, the corresponding paradigms suffer from their low-level. The purpose of this work-in-progress is to start from a (non-distributed) graph programming language, and translate its buildings blocks/constructs into lower level programming platforms. The graph programming language depends on graph rewriting, and its constructs are proved to cover any graph computable program.