Iterative Search for Reconfigurable Accelerator Blocks with a Compiler in the Loop

IEEE TCAD, October 2018
BibTeX
 @article{Willsey2018Iterative,
  author={
    Max Willsey and
    Vincent T. Lee and
    Alvin Cheung and
    Rastislav Bod{\'i}k and
    Luis Ceze
  },
  journal={IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems},
  title={Iterative Search for Reconfigurable Accelerator Blocks with a Compiler in the Loop},
  year={2018},
  volume={},
  number={},
  pages={1-1},
  abstract={
    Domain-specific reconfigurable accelerators (DSRAs) achieve high performance
    and energy efficiency by using specialized processing elements (PEs) instead
    of general-purpose alternatives. However, the process of designing, selecting,
    and refining the reconfigurable PEs that compose the accelerator fabric has
    remained a manual and difficult task. This paper presents Reconfigurable
    Accelerator Design using Iterative Search for Hardware (RADISH) which is a
    full-stack framework for automatically identifying and generating PEs from an
    application corpus. RADISH uses a genetic algorithm to iteratively search for
    and refine the proposed PEs with a compiler-in-the-loop to guide the search.
    We show that RADISH-generated PEs can generalize to both larger instances of
    the same application as well as other previously unseen applications within
    the same domain. We evaluate a CGRA architecture using our RADISH-generated
    PEs and show it achieves a geometric mean improvement of up to $2.14\times$
    and $2.4\times$ power and area respectively over an ALU-based CGRA designs. In
    terms of energy, our generated designs achieve a geometric mean improvement of
    $2.5\times$ but can achieve gains up to $28.9\times$.
  },
  keywords={Genetic algorithms;Tools;Hardware;Computer architecture;Python;Field programmable gate arrays;DSL},
  doi={10.1109/TCAD.2018.2878194},
  ISSN={0278-0070},
  month={10},
}
 

Abstract

Domain-specific reconfigurable accelerators (DSRAs) achieve high performance and energy efficiency by using specialized processing elements (PEs) instead of general-purpose alternatives. However, the process of designing, selecting, and refining the reconfigurable PEs that compose the accelerator fabric has remained a manual and difficult task. This paper presents Reconfigurable Accelerator Design using Iterative Search for Hardware (RADISH) which is a full-stack framework for automatically identifying and generating PEs from an application corpus. RADISH uses a genetic algorithm to iteratively search for and refine the proposed PEs with a compiler-in-the-loop to guide the search. We show that RADISH-generated PEs can generalize to both larger instances of the same application as well as other previously unseen applications within the same domain. We evaluate a CGRA architecture using our RADISH-generated PEs and show it achieves a geometric mean improvement of up to 2.14x and 2.4x power and area respectively over an ALU-based CGRA designs. In terms of energy, our generated designs achieve a geometric mean improvement of 2.5x but can achieve gains up to 28.9x.