Developments of Approximate Computing: From Algorithm Level to System Level
Authors:
- 1 January 2022
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 72, 52-56
Abstract
Low power is one of the critical parameters in the Very Large Scale Integration (VLSI) circuits design. In recent years, approximate computing techniques have improved the design metrics delay, area, and power or energy with a tradeoff inaccuracy. Computational intensive and error-tolerant applications are more suitable to adopt approximation Methods. Approximate computing is an emerging computing paradigm in low power VLSI design. It provides low power consumption at the cost of accuracy. This paper presents an introduction, applications, and developments in approximate computing.Keywords
This publication has 11 references indexed in Scilit:
- Low-Power Digital Signal Processing Using Approximate AddersIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2012
- Trading Accuracy for Power in a Multiplier ArchitectureJournal of Low Power Electronics, 2011
- Real-time H.264 video encoding in software with fast mode decision and dynamic complexity controlACM Transactions on Multimedia Computing, Communications, and Applications, 2010
- Computation Error Analysis in Digital Signal Processing Systems With Overscaled Supply VoltageIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2009
- Error-Resilient Motion Estimation ArchitectureIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2008
- Speeding up processing with approximation circuitsComputer, 2004
- Reliable and efficient system-on-chip designComputer, 2004
- A Voltage Overscaled Low-Power Digital Filter ICIEEE Journal of Solid-State Circuits, 2004
- Low-power filtering via adaptive error-cancellationIEEE Transactions on Signal Processing, 2003
- Soft digital signal processingIEEE Transactions on Very Large Scale Integration (VLSI) Systems, 2001