Abstract
As one of the most promising energy-efficient emerging paradigms for designing digital systems, approximate computing has attracted a significant attention in recent years. Applications utilizing approximate computing (AxC) can tolerate some loss of quality in the computed results for attaining high performance. Approximate arithmetic circuits have been extensively studied; however, their application at system level has not been extensively pursued. Furthermore, when approximate arithmetic circuits are applied at system level, error-accumulation effects and a convergence problem may occur in computation. Multiple approximate components can interact in a typical datapath, hence benefiting from each other. Many applications require more complex datapaths than a single multiplication. In this paper, a hardware/software co-design methodology for adaptive approximate computing is proposed. It makes use of feature constraints to guide the approximate computation at various accuracy levels in each iteration of the learning process in Artificial Neural Networks (ANNs). The proposed adaptive methodology also considers the input operand distribution and the hybrid approximation. Compared with a baseline design, the proposed method significantly reduces the power-delay product while incurring in only a small loss of accuracy. Simulation and a case study of image segmentation validate the effectiveness of the proposed methodology.
Highlights
As computer systems become pervasive, computing workloads have significantly increased due to new areas such as big data and IoT, the computing landscape has become more complex over the last decade
Semi-supervised learning is a type of supervised learning that uses unlabeled data to train a small amount of labeled data and a large amount of unlabeled data; k-means clustering divides n observations into k clusters, and each observation belongs to the nearest mean cluster
Larger p leads to simpler logic, but not always to less hardware; these results show that the dynamic power is dependent on the input operand distribution
Summary
As computer systems become pervasive, computing workloads have significantly increased due to new areas such as big data and IoT, the computing landscape has become more complex over the last decade. In [16], efficient approximate redundant binary multipliers for error-tolerant applications with high accuracy have been proposed; most of these works only consider approximate designs at circuit level, so error effects at system or algorithm level are not fully addressed. Approximate computing cannot be fully exploited by only considering hardware (circuits, architecture and memory) or software (application, algorithms and compile stack); the fundamental nature of machine learning (ML) workloads requires that the barriers between abstract layers to be clearly broken, so that an efficient ML system can be realized by AxC including cross-layer collaborative design. The goal is to approximate the mapping function, such that when a new input data (X ) is provided, the output variable (Y ) can be predicted These techniques are used in feedforward or multilayer perceptron (MLP) models.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
