05. GPU Approximation Acceleration For Scientific Applications
Authors: Ang Li (Pacific Northwest National Laboratory)Shuaiwen Leon Song (Pacific Northwest National Laboratory)
Abstract: Approximate computing, the technique that sacrifices certain amount of accuracy in exchange for substantial performance boost or power reduction, is one of the most promising solutions to enable power control and performance scaling towards exascale. In this poster, we introduce a transparent, tractable, and portable design framework for SFU-driven approximate acceleration on GPUs, providing fine-grained tuning for performance and accuracy trade-offs. Our design is software-based and requires no hardware or application modifications.
Two-page extended abstract: pdf