Self-Aware Processing for Adaptive Resource Optimization of Advanced Computing Systems

Impact Technologies, in collaboration with our partners Raytheon and Georgia Tech, is developing a suite of integrated components of a self-aware processing and adaptive, intelligent control system for next generation computing systems. During the Phase I program, the team developed a flexible architecture that supported development and implementation of organic computing modules. We integrated a suite of cognitive technologies to provide the constituent capabilities for self monitoring and reconfigurable computing, resulting in a self-aware framework that was shown to be capable of balancing system resources, and mission priorities to efficiently and dynamically meet goal driven requirements. The proposed Phase II program will focus on core innovations in three main areas. First, we will re-formulate the architecture and conceptual framework from Phase I into a more specific self-aware processing system prototype for polymorphous computing arrays (PCA) – specifically associated with the MONARCH platform. Second, working with our team member, we will enhance current simulation tools and generate new development tools for the FPCA framework that will enable implementation of the using high-level languages (such as C/C++). Third, we will work with our commercial partners to analyze how we can transition the self-aware concepts developed under Phase II to the exploding multi/many-core industry.