Fundamental Underpinnings of Reconfigurable Computing Architectures

Abstract

This paper introduces a unified framework for understanding advantages and tradeoffs in reconfigurable computing architectures.ABSTRACT | Reconfigurable architectures are a distinct point in the larger design space that includes programmable processors and nonprogrammable fixed-function devices. In this paper, we identify the major parameters that distinguish architectures in this design space and draw connections between these parameters and physical requirements (e.g., energy, delay, and area) and application characteristics (e.g., word width, locality). Building on these connections, we identify the fundamental advantages that reconfigurable architectures can offer.

Reconfigurable Computing Architectures

“…Instruction and data memory consume most of the energy in processors, while spatial architectures eliminate instruction energy and reduce data memory energy using smaller, distributed memories local to the computation. The companion paper shows how the locality exploitation in spatially distributed computations fundamentally allows them to achieve lower energy than sequential computations [24].…”

“…For the evolution of FPGAs for traditional usage, including the impact of technology, see the companion article [23]. For a theoretical and quantitative comparison of reconfigurable architectures, see [24].…”

“…Tabula now offers a commercial multicontext FPGA [51]. The companion paper [24] identifies conditions under which a multicontext FPGA can be lower energy than a single-context FPGA.…”

“…GPGPUs were initially focused on SIMD, single-precision, floating-point computing tasks, and do not support the spatially local communications typical of FPGAs and CGRAs. The companion paper [24] shows why locality exploitation can be a key advantage for reconfigurable architectures. For compiled CUDA, the native language of many GPGPUs, the GPGPUs provide 1-4Â higher throughput than FPGAs, but at a cost of over 4-16Â the energy per operation [60].…”

A Survey of Coarse-Grained Reconfigurable Architecture and Design

“…In recent years, many surveys and books about reconfigurable computing have been published [15,[34][35][36][37][38][39][40][41][42][43][44][45]. Some of them concentrate on FPGAs and pay less attention to CGRAs [35,37,40,[42][43][44][45]. The literature focused on CGRAs has covered a lot of topics, including classifications, architectures, applications, compilation methods and tools, and challenges.…”