Don't Let RPCs Constrain Your API

Bittman, Daniel; Soulé, Robert; Miller, Ethan L.; Shrivastav, Vishal; Mehra, Pankaj; Boisvert, Matthew; Silberschatz, Abraham; Alvaro, Peter

doi:10.1145/3484266.3487389

Cited by 4 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While there are hundreds of commercial accelerator designs [1,[3][4][5][6]22], the API calls used to program them lack the portability and flexibility of software libraries [10,53] making offloading the domain of experts [107]. Manually migrating to new software APIs is complex and timeconsuming [46,72], and made more challenging by the inability for APIs to hide the complex eccentricities exposed by real hardware [25,26].…”

Section: Introductionmentioning

confidence: 99%

Bind the gap: compiling real software to hardware FFT accelerators

Woodruff

Armengol-Estapé

Ainsworth

et al. 2022

Proceedings of the 43rd ACM SIGPLAN International Conference on Programming Language Design and Implementation

View full text Add to dashboard Cite

Specialized hardware accelerators continue to be a source of performance improvement. However, such specialization comes at a programming price. The fundamental issue is that of a mismatch between the diversity of user code and the functionality of fixed hardware, limiting its wider uptake.Here we focus on a particular set of accelerators: those for Fast Fourier Transforms. We present FACC (Fourier ACcelerator Compiler), a novel approach to automatically map legacy code to Fourier Transform accelerators. It automatically generates drop-in replacement adapters using Input-Output (IO)-based program synthesis that bridge the gap between user code and accelerators. We apply FACC to unmodified GitHub C programs of varying complexity and compare against two existing approaches. We target FACC to a high-performance library, FFTW, and two hardware accelerators, the NXP PowerQuad and the Analog Devices FFTA, and demonstrate mean speedups of 9x, 17x and 27x respectively.

show abstract