With semiconductor scaling reaching physical limits, performance and power consumption are ever more critical aspects in the design of emerging computer systems. Next-generation systems ranging from small embedded devices to city-scale data centers are expected to contain a heterogeneous mix of multiple CPUs, GPUs, FPGAs, accelerators and memory stacks all in the same package or die. This poses fundamental new challenges for designing, programming and managing of such systems. Fast and accurate power and performance models of architectures and the applications running on them are essential for evaluating design options before systems are built. With increase in complexity of systems and applications, traditional simulation-based or analytical modeling approaches are rapidly becoming too slow or inaccurate to be feasible.
This project aims to investigate use of advanced machine learning-based, predictive methodologies to rapidly estimate the performance and power consumption of future generation products at early design stages using observations obtained on commercially available silicon today, specifically to aid in heterogeneous system design, programming and runtime management. Such techniques will allow efficient design cycles ensuring that the next-generation computing infrastructure meets the consumer and society's needs. In addition, the project will be complemented by educational, outreach and active industry collaboration and technology transfer activities.