Tuesday, November 13, 2012

Approximate Computing 2

A reading list for Approximate Computing /Soft Computing/ Imprecise Computing, Stochastic Computation

A NumericalOptimization-based Methodology forApplication Robustification: Transforming Applicationsfor Error Tolerance

Joseph Sloan, John Sartori, and Rakesh Kumar. " Exploiting Application-Level Error Tolerance in Software Design for Stochastic Processors. " In the 49th Design and Automation Conference. DAC, San Francisco, June 2012. (PDF). (invited)

Best-effort semantic document search on GPUs

Sunday, November 4, 2012

Database of integer sequences

http://oeis.org/

Approximate Computing

A reading list for Approximate Computing /Soft Computing/ Imprecise Computing, Stochastic Computation

Joseph Bates: http://web.media.mit.edu/~bates/Welcome_files/BatesPosterOct2010.pdf
http://web.media.mit.edu/~bates/Summary_files/BatesTalk.pdf

http://en.wikipedia.org/wiki/Bate%27s_chip

Doug Burger - Microsoft
http://research.ihost.com/weed2012/pdfs/paper%20G.pdf
http://sampa.cs.washington.edu/public/uploads/b/bc/Npu-dasi12.pdf
http://www.cs.utexas.edu/~hadi/doc/paper/2012-asplos-truffle.pdf
http://homes.cs.washington.edu/~asampson/media/papers/enerj-pldi2011.pdf

CMU
http://www.cs.cmu.edu/~gpekhime/Projects/15740/paper.pdf

IBM:
http://www.research.cornell.edu/KIC/events/Computing2008/Nair_Cornell_08.pdf


Darpa
http://csl.stanford.edu/~christos/publications/2012.isat.ACSWTP.slides.pdf

University of Illinois at Urbana-Champaign - Naresh R. Shanbhag
http://passat.crhc.illinois.edu/rakeshk/dac_10_invited_cam.pdf
ANT
http://www.eurasip.org/Proceedings/Eusipco/Eusipco2000/SESSIONS/THUPM/SS3/CR1825.PDF
Stochastic Computing: Embracing Errors in Architecture and Design of Processors and Applications

Energy Aware Probabilistic Multiplier
Energy-aware probabilistic multiplier: design and analysis, MSK Lau et al, Proceedings of
the 2009 international conference on Compilers, architecture, and synthesis for embedded
systems

Power Benefits of Imprecise ComputerArithmetic

IMPACT: IMPrecise adders for low-power approximate computing 

Interesting Summary from above http://csl.stanford.edu/~christos/publications/2012.isat.ACSWTP.slides.pdf
1. Exact output with approximate HW
2. Approximate output with deterministic HW (unsound SW transformations)
3. Approximate output with approximate HW

An interesting keypoint:
statistical nature of application-level has to match statistical nature of underlying device!


Survey of Stochastic Computing
Book -Stochastic Computing
http://www.nowpublishers.com/product.aspx?product=EDA&doi=1000000021&section=xstart

Motivation for Neuromorphic computing systems in stochastic computing