This is an introductory paper of fresh started research that concerns the development of new HPC parallelization strategies based on intrinsically parallel unconventional computation paradigms. The main goal of this research is to use theses developed parallelization technologies for overcoming modern HPC energy consumption issues. Development of new parallelization technologies will be focused on a particular type of bio-inspired models of computation: the membrane computing paradigm. The solutions for the set of selected problems will be obtained by application of membrane computing. After that these solutions will be decomposed into elementary steps and HPC patterns will be created for each step using ready libraries in order to implement proposed solution on HPC efficiently. A pattern here means an algorithmic skeleton corresponding to a physical hardware architecture as well as the necessary implementation libraries. Basing on the feedback of HPC patterns development, the membrane models would be eventually adjusted in order to obtain simpler and/or more efficient patterns.