A Design Methodology for Domain-Optimized Power-Efficient Supercomputing

TitleA Design Methodology for Domain-Optimized Power-Efficient Supercomputing
Publication TypeConference Paper
Year of Publication2009
AuthorsMohiyuddin, M., Murphy M., Oliker L., Shalf J., Wawrzynek J., & Williams S.
Conference NameSupercomputing '09
Date Published11/2009
Conference LocationPortlad, Oregon

As power has become the pre-eminent design constraint for future HPC systems, computational efficiency is being emphasized over simply peak performance. Recently, static benchmark codes have been used to find a power efficient architecture. Unfortunately, because compilers generate sub-optimal code, benchmark performance can be a poor indicator of the performance potential of architecture design points. Therefore, we present hardware/software co-tuning as a novel approach for system design, in which traditional architecture space exploration is tightly coupled with software auto-tuning for delivering substantial improvements in area and power efficiency. We demonstrate the proposed methodology by exploring the parameter space of a Tensilica-based multi-processor running three of the most heavily used kernels in scientific computing, each with widely varying micro-architectural requirements: sparse matrix vector multiplication, stencil-based computations, and general matrix-matrix multiplication. Results demonstrate that co-tuning significantly improves hardware area and energy efficiency -- a key driver for next generation of HPC system design.