Parsimonious circuits for error-tolerant applications through probabilistic logic minimization

Authors:
Avinash Lingamneni;Christian Enz;Krishna Palem;Christian Piguet
Affiliations:
NTU-Rice Institute for Sustainable and Applied Infodynamics, Department of ECE, Rice University, Houston and Integrated and Wireless Systems Division, Centre Suisse d'Electronique et de Microtechn ...;Integrated and Wireless Systems Division, Centre Suisse d'Electronique et de Microtechnique SA, Neuchatel, Switzerland;NTU-Rice Institute for Sustainable and Applied Infodynamics, Department of ECE, Rice University, Houston;Integrated and Wireless Systems Division, Centre Suisse d'Electronique et de Microtechnique SA, Neuchatel, Switzerland
Venue:
PATMOS'11 Proceedings of the 21st international conference on Integrated circuit and system design: power and timing modeling, optimization, and simulation
Year:
2011

Citing 10
Cited 5

Energy aware algorithm design via probabilistic computing: from algorithms and models to Moore's law and novel (semiconductor) devices

Proceedings of the 2003 international conference on Compilers, architecture and synthesis for embedded systems
Energy-efficient motion estimation using error-tolerance

Proceedings of the 2006 international symposium on Low power electronics and design
Probabilistic arithmetic and energy efficient embedded signal processing

CASES '06 Proceedings of the 2006 international conference on Compilers, architecture and synthesis for embedded systems
Approximate logic circuits for low overhead, non-intrusive concurrent error detection

Proceedings of the conference on Design, automation and test in Europe
Highly energy and performance efficient embedded computing through approximately correct arithmetic: a mathematical foundation and preliminary experimental validation

CASES '08 Proceedings of the 2008 international conference on Compilers, architectures and synthesis for embedded systems
Impact of supply voltage variations on full adder delay: analysis and comparison

IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Significance driven computation: a voltage-scalable, variation-aware, quality-tuning motion estimator

Proceedings of the 14th ACM/IEEE international symposium on Low power electronics and design
Sustaining moore's law in embedded computing through probabilistic and approximate design: retrospects and prospects

CASES '09 Proceedings of the 2009 international conference on Compilers, architecture, and synthesis for embedded systems
Scalable stochastic processors

Proceedings of the Conference on Design, Automation and Test in Europe
Approximate logic synthesis for error tolerant applications

Proceedings of the Conference on Design, Automation and Test in Europe

Algorithmic methodologies for ultra-efficient inexact architectures for sustaining technology scaling

Proceedings of the 9th conference on Computing Frontiers
What to do about the end of Moore's law, probably!

Proceedings of the 49th Annual Design Automation Conference
Ten Years of Building Broken Chips: The Physics and Engineering of Inexact Computing

ACM Transactions on Embedded Computing Systems (TECS) - Special Section on Probabilistic Embedded Computing
Synthesizing Parsimonious Inexact Circuits through Probabilistic Design Techniques

ACM Transactions on Embedded Computing Systems (TECS) - Special Section on Probabilistic Embedded Computing
Improving energy gains of inexact DSP hardware through reciprocative error compensation

Proceedings of the 50th Annual Design Automation Conference

Quantified Score

Hi-index	0.00

Visualization

Abstract

Contrary to the existing techniques to realize inexact circuits that relied mostly on scaling of supply voltage or pruning of "least-significant" components in conventional correct circuits to achieve cost (energy, delay and/or area) and accuracy tradeoffs, we propose a novel technique called Probabilistic Logic Minimization which relies on synthesizing an inexact circuit in the first place resulting in zero hardware overhead. Extensive simulations of the datapath elements designed using the proposed technique demonstrate that normalized gains as high as 2X- 9.5X in the Energy-Delay-Area product can be obtained when compared to the corresponding correct designs, with a relative error magnitude percentage as low as 0.001% upto 1%.