%0 Journal Article %J IEEE Computer Architecture Letters %D 2010 %T The Accelerator Store framework for high-performance, low-power accelerator-based systems %A Michael Lyons %A Mark Hempstead %A Gu Wei %A David Brooks %K accelerator-store %X Hardware acceleration can increase performance and reduce energy consumption. To maximize these benefits, accelerator- based systems that emphasize computation on accelerators (rather than on general purpose cores) should be used. We introduce the “accelerator store,” a structure for sharing memory between accelerators in these accelerator-based systems. The accelerator store simplifies accelerator I/O and reduces area by mapping memory to accelerators when needed at runtime. Preliminary results demonstrate a 30% system area reduction with no energy overhead and less than 1% performance overhead in contrast to conventional DMA schemes. %B IEEE Computer Architecture Letters %I IEEE %V 9 %P 53-56 %G eng %U https://doi.org/10.1109/L-CA.2010.16 %N 2