Optical implementation of probabilistic graphical models

Pierre Alexandre Blanche; Masoud Babaeian; Madeleine Glick; John Wissinger; Robert Norwood; Nasser Peyghambarian; Mark Neifeld; Ratchaneekorn Thamvichai

doi:10.1109/ICRC.2016.7738702

Optical implementation of probabilistic graphical models

Pierre Alexandre Blanche, Masoud Babaeian, Madeleine Glick, John Wissinger, Robert Norwood, Nasser Peyghambarian, Mark Neifeld, Ratchaneekorn Thamvichai

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

7 Scopus citations

Abstract

We are investigating the use of optics to solve highly connected graphical models by probabilistic inference, and more specifically the sum-product message passing algorithm. We are examining the fundamental limit of size and power requirement according to the best multiplexing strategy we have found. For a million nodes, and an alphabet of a hundred, we found that the minimum size for the optical implementation is 10mm³, and the lowest bound for the power is 200 watts for operation at the shot noise limit. The various functions required for the algorithm to be operational are presented and potential implementations are discussed. These include a vector matrix multiplication using spectral hole burning, a logarithm carried out with two photon absorption, an exponential performed with saturable absorption, a normalization executed with an thermo-optics interferometer, and a wavelength remapping accomplished with a pump-probe amplifier.

Original language	English (US)
Title of host publication	2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781509013708
DOIs	https://doi.org/10.1109/ICRC.2016.7738702
State	Published - Nov 8 2016
Event	2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - San Diego, United States Duration: Oct 17 2016 → Oct 19 2016

Publication series

Name	2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings

Other

Other	2016 IEEE International Conference on Rebooting Computing, ICRC 2016
Country/Territory	United States
City	San Diego
Period	10/17/16 → 10/19/16

ASJC Scopus subject areas

Hardware and Architecture

Access to Document

10.1109/ICRC.2016.7738702

Cite this

Blanche, P. A., Babaeian, M., Glick, M., Wissinger, J., Norwood, R., Peyghambarian, N., Neifeld, M., & Thamvichai, R. (2016). Optical implementation of probabilistic graphical models. In 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings Article 7738702 (2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICRC.2016.7738702

Optical implementation of probabilistic graphical models. / Blanche, Pierre Alexandre; Babaeian, Masoud; Glick, Madeleine et al.
2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2016. 7738702 (2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Blanche, PA, Babaeian, M, Glick, M, Wissinger, J, Norwood, R , Peyghambarian, N, Neifeld, M & Thamvichai, R 2016, Optical implementation of probabilistic graphical models. in 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings., 7738702, 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings, Institute of Electrical and Electronics Engineers Inc., 2016 IEEE International Conference on Rebooting Computing, ICRC 2016, San Diego, United States, 10/17/16. https://doi.org/10.1109/ICRC.2016.7738702

Blanche PA, Babaeian M, Glick M, Wissinger J, Norwood R , Peyghambarian N et al. Optical implementation of probabilistic graphical models. In 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc. 2016. 7738702. (2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings). doi: 10.1109/ICRC.2016.7738702

Blanche, Pierre Alexandre ; Babaeian, Masoud ; Glick, Madeleine et al. / Optical implementation of probabilistic graphical models. 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2016. (2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings).

@inproceedings{b28de2a2593a454db8a46fe784536a50,

title = "Optical implementation of probabilistic graphical models",

abstract = "We are investigating the use of optics to solve highly connected graphical models by probabilistic inference, and more specifically the sum-product message passing algorithm. We are examining the fundamental limit of size and power requirement according to the best multiplexing strategy we have found. For a million nodes, and an alphabet of a hundred, we found that the minimum size for the optical implementation is 10mm3, and the lowest bound for the power is 200 watts for operation at the shot noise limit. The various functions required for the algorithm to be operational are presented and potential implementations are discussed. These include a vector matrix multiplication using spectral hole burning, a logarithm carried out with two photon absorption, an exponential performed with saturable absorption, a normalization executed with an thermo-optics interferometer, and a wavelength remapping accomplished with a pump-probe amplifier.",

author = "Blanche, {Pierre Alexandre} and Masoud Babaeian and Madeleine Glick and John Wissinger and Robert Norwood and Nasser Peyghambarian and Mark Neifeld and Ratchaneekorn Thamvichai",

note = "Publisher Copyright: {\textcopyright} 2016 IEEE.; 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 ; Conference date: 17-10-2016 Through 19-10-2016",

year = "2016",

month = nov,

day = "8",

doi = "10.1109/ICRC.2016.7738702",

language = "English (US)",

series = "2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings",

}

TY - GEN

T1 - Optical implementation of probabilistic graphical models

AU - Blanche, Pierre Alexandre

AU - Babaeian, Masoud

AU - Glick, Madeleine

AU - Wissinger, John

AU - Norwood, Robert

AU - Peyghambarian, Nasser

AU - Neifeld, Mark

AU - Thamvichai, Ratchaneekorn

PY - 2016/11/8

Y1 - 2016/11/8

N2 - We are investigating the use of optics to solve highly connected graphical models by probabilistic inference, and more specifically the sum-product message passing algorithm. We are examining the fundamental limit of size and power requirement according to the best multiplexing strategy we have found. For a million nodes, and an alphabet of a hundred, we found that the minimum size for the optical implementation is 10mm3, and the lowest bound for the power is 200 watts for operation at the shot noise limit. The various functions required for the algorithm to be operational are presented and potential implementations are discussed. These include a vector matrix multiplication using spectral hole burning, a logarithm carried out with two photon absorption, an exponential performed with saturable absorption, a normalization executed with an thermo-optics interferometer, and a wavelength remapping accomplished with a pump-probe amplifier.

AB - We are investigating the use of optics to solve highly connected graphical models by probabilistic inference, and more specifically the sum-product message passing algorithm. We are examining the fundamental limit of size and power requirement according to the best multiplexing strategy we have found. For a million nodes, and an alphabet of a hundred, we found that the minimum size for the optical implementation is 10mm3, and the lowest bound for the power is 200 watts for operation at the shot noise limit. The various functions required for the algorithm to be operational are presented and potential implementations are discussed. These include a vector matrix multiplication using spectral hole burning, a logarithm carried out with two photon absorption, an exponential performed with saturable absorption, a normalization executed with an thermo-optics interferometer, and a wavelength remapping accomplished with a pump-probe amplifier.

UR - http://www.scopus.com/inward/record.url?scp=85006001656&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006001656&partnerID=8YFLogxK

U2 - 10.1109/ICRC.2016.7738702

DO - 10.1109/ICRC.2016.7738702

M3 - Conference contribution

AN - SCOPUS:85006001656

T3 - 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings

BT - 2016 IEEE International Conference on Rebooting Computing, ICRC 2016 - Conference Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2016 IEEE International Conference on Rebooting Computing, ICRC 2016

Y2 - 17 October 2016 through 19 October 2016

ER -

Optical implementation of probabilistic graphical models

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this