Aging Aware Approximation to Reduce Power Consumption
Loading...
Links to Files
Permanent Link
Author/Creator
Author/Creator ORCID
Date
2021-01-01
Type of Work
Department
Computer Science and Electrical Engineering
Program
Engineering, Computer
Citation of Original Publication
Rights
This item may be protected under Title 17 of the U.S. Copyright Law. It is made available by UMBC for non-commercial research and education. For permission to publish or reproduce, please see http://aok.lib.umbc.edu/specoll/repro.php or contact Special Collections at speccoll(at)umbc.edu
Access limited to the UMBC community. Item may possibly be obtained via Interlibrary Loan through a local library, pending author/copyright holder's permission.
Access limited to the UMBC community. Item may possibly be obtained via Interlibrary Loan thorugh a local library, pending author/copyright holder's permission.
Access limited to the UMBC community. Item may possibly be obtained via Interlibrary Loan through a local library, pending author/copyright holder's permission.
Access limited to the UMBC community. Item may possibly be obtained via Interlibrary Loan thorugh a local library, pending author/copyright holder's permission.
Abstract
Over the past decade, there has been an increased desire to produce low-power consumption devices. A popular approach to this problem is the use of approximate computing. Recent methodologies to achieve power saving embraces the error tolerance nature of certain applications such as image compression/decompression applications for which exploiting their innate approximation can result in a significant increase in the circuits' energy efficiency with only a minor increase to the circuits' error rate. While different schemes have been presented in the literature for increasing energy efficiency, most of them fail to consider the efficiency of the proposed schemes when the circuit is aged. This research will review previous approximation-induced power reduction algorithms and present an algorithm for reducing power consumption that is still efficient when the circuit is aged, i.e., providing long-lasting reliability for the approximated circuit.To show the efficiency of the proposed scheme, we utilized the Discrete Cosine Transform (DCT) circuit due to its nature of being approximable. Three different approximation methods were implemented to compare both the power reduction and the error rate occurs due to the approximation over the course of circuit aging. Each method selects a subset of the target circuit's gates, based on different criteria, to be operated under near-threshold condition to decrease the power consumption. The experimental results show that the proposed method can reduce the power consumption by around 13% while diminishes the error rates compared to the previous methods over the course of 1 year of usage.