High-level synthesis of approximate hardware under joint precision and voltage scaling
High-level synthesis of approximate hardware under joint precision and voltage scaling
复制标题
联合精度和电压缩放下近似硬件的高级综合
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
A. Gerstlauer
中科院分区：
文献类型：
--
作者：
Seogoo Lee;L. John;A. Gerstlauer
近年来，近似计算已成为一种有前途的方法，可以在计算输出的质量和节能之间进行权衡。在本文中，我们提出了一种近似高级综合 (AHLS) 方法，该方法根据准确的高级 C 描述输出质量能量优化的寄存器传输级实现。现有的 AHLS 工作仅考虑硬件近似下的节能切换活动。相比之下，我们的目标是提供一种通用的 AHLS 解决方案，在减少处理时间的情况下，该解决方案还考虑电压缩放。为了最大限度地降低电压和相关的能耗，我们将通过位舍入进行的操作级别近似和更积极的操作消除作为近似技术。在这种近似下最佳地利用扩展机会需要与调度任务紧密交互。我们通过将估计近似调度影响的优化过程与快速而准确的质量能量模型和高效的优化求解器相结合来解决这个问题，以建设性地找到接近最优的解决方案。结果表明，在考虑电压调节时，与仅考虑开关活动的方法相比，可以实现高达 24.5% 的节能效果。与详尽搜索相比，我们的启发式求解器能够在平均节能 0.1% 以内找到解决方案，同时比基于模拟的方法快 1,400 χ。
In recent years, approximate computing has emerged as a promising approach to trade off quality of computed outputs for energy savings. In this paper, we present an approximate high-level synthesis (AHLS) approach that outputs a quality-energy optimized register-transfer-level implementation from an accurate high-level C description. Existing AHLS work only considers switching activity for energy savings under hardware approximations. By contrast, we aim to provide a general AHLS solution that also considers voltage scaling given a reduced processing time. To maximize voltage and associated energy reductions, we include both operation-level approximations by bit rounding and more aggressive operation eliminations as approximation techniques. Optimally exploiting scaling opportunities under such approximations requires tight interaction with scheduling tasks. We address this problem by combining an optimization pass that estimates the scheduling impact of approximations with fast yet accurate quality-energy models and an efficient optimization solver to find near-optimal solutions constructively. Results show that when considering voltage scaling, up to 24.5% higher energy savings can be achieved compared to approaches that only consider switching activity. Our heuristic solver is able to find solutions within 0.1% of average energy savings compared to an exhaustive search, all while being up to 1,400 χ faster than simulation-based methods.