Cross-Architectural Study of Custom Reconfigurable Devices Using Crowdsourcing

Coarse grained reconfigurable architectures (CGRAs) are promising due to the ability to highly customize such architectures to an application domain. However, good tools and good algorithms to map benchmarks onto these architectures are needed to support design space exploration for CGRAs. In particular, the mapping problem has been difficult to solve in a satisfying and general way. In this paper, we present an architectural design flow using crowd sourcing to provide mappings of benchmarks onto new architectures. We show that the crowd can provide high quality, reliable mappings, outperforming our custom Simulated Annealing algorithm in 37 of 42 trials. We further show that the crowd can provide other types of feedback that are difficult to obtain from an automatic mapping algorithm. Our proof of concept cross-architectural study concludes that a mesh architecture with 8Way connectivity outperforms the other interconnection options tested. A stripe architecture with dedicated vertical routes (StripeDR) performs competitively as well.