Applications and Techniques for Fast Machine Learning in Science

Allison Mc Carn Deiana; Nhan Tran; Joshua Agar; Michaela Blott; Giuseppe Di Guglielmo; Javier Duarte; Philip Harris; Scott Hauck; Mia Liu; Mark S. Neubauer; Jennifer Ngadiuba; Seda Ogrenci-Memik; Maurizio Pierini; Thea Aarrestad; Steffen Bähr; Jürgen Becker; Anne Sophie Berthold; Richard J. Bonventre; Tomás E. Müller Bravo; Markus Diefenthaler; Zhen Dong; Nick Fritzsche; Amir Gholami; Ekaterina Govorkova; Dongning Guo; Kyle J. Hazelwood; Christian Herwig; Babar Khan; Sehoon Kim; Thomas Klijnsma; Yaling Liu; Kin Ho Lo; Tri Nguyen; Gianantonio Pezzullo; Seyedramin Rasoulinezhad; Ryan A. Rivera; Kate Scholberg; Justin Selig; Sougata Sen; Dmitri Strukov; William Tang; Savannah Thais; Kai Lukas Unger; Ricardo Vilalta; Belina von Krosigk; Shen Wang; Thomas K. Warburton

doi:10.3389/fdata.2022.787421

Applications and Techniques for Fast Machine Learning in Science

Allison Mc Carn Deiana^*, Nhan Tran^*, Joshua Agar, Michaela Blott, Giuseppe Di Guglielmo, Javier Duarte, Philip Harris, Scott Hauck, Mia Liu, Mark S. Neubauer, Jennifer Ngadiuba, Seda Ogrenci-Memik, Maurizio Pierini, Thea Aarrestad, Steffen Bähr, Jürgen Becker, Anne Sophie Berthold, Richard J. Bonventre, Tomás E. Müller Bravo, Markus DiefenthalerZhen Dong, Nick Fritzsche, Amir Gholami, Ekaterina Govorkova, Dongning Guo, Kyle J. Hazelwood, Christian Herwig, Babar Khan, Sehoon Kim, Thomas Klijnsma, Yaling Liu, Kin Ho Lo, Tri Nguyen, Gianantonio Pezzullo, Seyedramin Rasoulinezhad, Ryan A. Rivera, Kate Scholberg, Justin Selig, Sougata Sen, Dmitri Strukov, William Tang, Savannah Thais, Kai Lukas Unger, Ricardo Vilalta, Belina von Krosigk, Shen Wang, Thomas K. Warburton

^*Corresponding author for this work

Electrical and Computer Engineering

Research output: Contribution to journal › Review article › peer-review

39 Scopus citations

Abstract

In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

Original language	English (US)
Article number	787421
Journal	Frontiers in Big Data
Volume	5
DOIs	https://doi.org/10.3389/fdata.2022.787421
State	Published - Apr 12 2022

Funding

We acknowledge the Fast Machine Learning collective as an open community of multi-domain experts and collaborators. This community was important for the development of this project. The work by AD was supported by the U.S. Department of Energy (DOE), Office of Science, Office of High Energy Physics, under Award No. DE-SC0010129, and the Fast Machine Learning in Science Workshop was financially supported by Southern Methodist University. The work by NT was supported by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the DOE, Office of Science, Office of High Energy Physics and the DOE Early Career Research program under Award No. DE-0000247070. The work by DS was supported by NSF E2CDA grant #1740352. JA acknowledges primary support from our DOE program and secondary support from National Science Foundation under grant TRIPODS + X: RES-1839234Y. The work of DG was supported in part by the National Science Foundation under Grant No. CNS-2003098 and by a gift from Intel Incorporation. YL acknowledges the support of this work from the National Institutes of Health grant of R01HL131750, and National Science Foundation grant of CBET 2039310. The work by MN was supported by the U.S. National Science Foundation under Cooperative Agreement OAC-1836650 and Award No. OAC-1934757. KS is supported by the U.S. Department of Energy and the National Science Foundation. The work by BK is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under the Emmy Noether Grant No. 420484612. The work by MD was supported by Jefferson Science Associates, LLC under Contract No. DE-AC05-06OR23177 with the DOE, Office of Science, Office of Nuclear Physics. We would like to acknowledge community members who have explicitly supported this work: Maria Acosta Flechas (Fermilab), Anthony Aportela (UC San Diego), Thomas Calvet (CPP Marseille), Leonardo Cristella (CERN), Daniel Diaz (UC San Diego), Caterina Doglioni (Lund), Maria Domenica Galati (University of Groningen), Elham E Khoda (University of Washington), Farah Fahim (Fermilab), Davide Giri (Columbia University), Benjamin Hawks (Fermilab), Duc Hoang (MIT), Burt Holzman (Fermilab), Shih-Chieh Hsu (University of Washington), Sergo Jindariani (Fermilab), Iris Johnson (Fermilab), Raghav Kansal (UC San Diego), Ryan Kastner (UC San Diego), Erik Katsavounidis (MIT), Jeffrey Krupa (MIT), Pan Li (Purdue University), Vladimir Loncar (CERN, Institute of Physics Belgrad), Sandeep Madireddy (ANL), Ethan Marx (MIT), Patrick McCormack (MIT) Andres Meza (UC San Diego), Jovan Mitrevski (Fermilab), Mohammed Attia Mohammed (CHEP-FU), Farouk Mokhtar (UC San Diego), Eric Moreno (MIT), Srishti Nagu (Lucknow University), Rohin Narayan (SMU), Noah Paladino (MIT), Adrian Alan Pol (CERN), Zhiqiang Que (Imperial College), Sang Eon Park (MIT), Subramanian Ramamoorthy 28, Dylan Rankin (MIT), Simon Rothman (MIT), Ashish Sharma (IIT Madras), Sioni Summers (CERN), Pietro Vischia (UC Louvain), Jean-Roch Vlimant (Caltech), Olivia Weng (UC San Diego). We acknowledge the Fast Machine Learning collective as an open community of multi-domain experts and collaborators. This community was important for the development of this project. The work by AD was supported by the U.S. Department of Energy (DOE), Office of Science, Office of High Energy Physics, under Award No. DE-SC0010129, and the Fast Machine Learning in Science Workshop was financially supported by Southern Methodist University. The work by NT was supported by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the DOE, Office of Science, Office of High Energy Physics and the DOE Early Career Research program under Award No. DE-0000247070. The work by DS was supported by NSF E2CDA grant #1740352. JA acknowledges primary support from our DOE program and secondary support from National Science Foundation under grant TRIPODS + X: RES-1839234Y. The work of DG was supported in part by the National Science Foundation under Grant No. CNS-2003098 and by a gift from Intel Incorporation. YL acknowledges the support of this work from the National Institutes of Health grant of R01HL131750, and National Science Foundation grant of CBET 2039310. The work by MN was supported by the U.S. National Science Foundation under Cooperative Agreement OAC-1836650 and Award No. OAC-1934757. KS is supported by the U.S. Department of Energy and the National Science Foundation. The work by BK is funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under the Emmy Noether Grant No. 420484612. The work by MD was supported by Jefferson Science Associates, LLC under Contract No. DE-AC05-06OR23177 with the DOE, Office of Science, Office of Nuclear Physics. We would like to acknowledge community members who have explicitly supported this work: Maria Acosta Flechas (Fermilab), Anthony Aportela (UC San Diego), Thomas Calvet (CPP Marseille), Leonardo Cristella (CERN), Daniel Diaz (UC San Diego), Caterina Doglioni (Lund), Maria Domenica Galati (University of Groningen), Elham E Khoda (University of Washington), Farah Fahim (Fermilab), Davide Giri (Columbia University), Benjamin Hawks (Fermilab), Duc Hoang (MIT), Burt Holzman (Fermilab), Shih-Chieh Hsu (University of Washington), Sergo Jindariani (Fermilab), Iris Johnson (Fermilab), Raghav Kansal (UC San Diego), Ryan Kastner (UC San Diego), Erik Katsavounidis (MIT), Jeffrey Krupa (MIT), Pan Li (Purdue University), Vladimir Loncar (CERN, Institute of Physics Belgrad), Sandeep Madireddy (ANL), Ethan Marx (MIT), Patrick McCormack (MIT) Andres Meza (UC San Diego), Jovan Mitrevski (Fermilab), Mohammed Attia Mohammed (CHEP-FU), Farouk Mokhtar (UC San Diego), Eric Moreno (MIT), Srishti Nagu (Lucknow University), Rohin Narayan (SMU), Noah Paladino (MIT), Adrian Alan Pol (CERN), Zhiqiang Que (Imperial College), Sang Eon Park (MIT), Subramanian Ramamoorthy , Dylan Rankin (MIT), Simon Rothman (MIT), Ashish Sharma (IIT Madras), Sioni Summers (CERN), Pietro Vischia (UC Louvain), Jean-Roch Vlimant (Caltech), Olivia Weng (UC San Diego). 28

Keywords

big data
codesign
coprocessors
fast machine learning
heterogeneous computing
machine learning for science
particle physics

ASJC Scopus subject areas

Computer Science (miscellaneous)
Information Systems
Artificial Intelligence

Access to Document

10.3389/fdata.2022.787421

Cite this

Deiana, A. M. C., Tran, N., Agar, J., Blott, M., Di Guglielmo, G., Duarte, J., Harris, P., Hauck, S., Liu, M., Neubauer, M. S., Ngadiuba, J., Ogrenci-Memik, S., Pierini, M., Aarrestad, T., Bähr, S., Becker, J., Berthold, A. S., Bonventre, R. J., Müller Bravo, T. E., ... Warburton, T. K. (2022). Applications and Techniques for Fast Machine Learning in Science. Frontiers in Big Data, 5, Article 787421. https://doi.org/10.3389/fdata.2022.787421

@article{a51d580dc58c4035893ece266ec7ee88,

title = "Applications and Techniques for Fast Machine Learning in Science",

abstract = "In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.",

keywords = "big data, codesign, coprocessors, fast machine learning, heterogeneous computing, machine learning for science, particle physics",

author = "Deiana, {Allison Mc Carn} and Nhan Tran and Joshua Agar and Michaela Blott and {Di Guglielmo}, Giuseppe and Javier Duarte and Philip Harris and Scott Hauck and Mia Liu and Neubauer, {Mark S.} and Jennifer Ngadiuba and Seda Ogrenci-Memik and Maurizio Pierini and Thea Aarrestad and Steffen B{\"a}hr and J{\"u}rgen Becker and Berthold, {Anne Sophie} and Bonventre, {Richard J.} and {M{\"u}ller Bravo}, {Tom{\'a}s E.} and Markus Diefenthaler and Zhen Dong and Nick Fritzsche and Amir Gholami and Ekaterina Govorkova and Dongning Guo and Hazelwood, {Kyle J.} and Christian Herwig and Babar Khan and Sehoon Kim and Thomas Klijnsma and Yaling Liu and Lo, {Kin Ho} and Tri Nguyen and Gianantonio Pezzullo and Seyedramin Rasoulinezhad and Rivera, {Ryan A.} and Kate Scholberg and Justin Selig and Sougata Sen and Dmitri Strukov and William Tang and Savannah Thais and Unger, {Kai Lukas} and Ricardo Vilalta and {von Krosigk}, Belina and Shen Wang and Warburton, {Thomas K.}",

note = "Publisher Copyright: Copyright {\textcopyright} 2022 Deiana, Tran, Agar, Blott, Di Guglielmo, Duarte, Harris, Hauck, Liu, Neubauer, Ngadiuba, Ogrenci-Memik, Pierini, Aarrestad, B{\"a}hr, Becker, Berthold, Bonventre, M{\"u}ller Bravo, Diefenthaler, Dong, Fritzsche, Gholami, Govorkova, Guo, Hazelwood, Herwig, Khan, Kim, Klijnsma, Liu, Lo, Nguyen, Pezzullo, Rasoulinezhad, Rivera, Scholberg, Selig, Sen, Strukov, Tang, Thais, Unger, Vilalta, von Krosigk, Wang and Warburton.",

year = "2022",

month = apr,

day = "12",

doi = "10.3389/fdata.2022.787421",

language = "English (US)",

volume = "5",

journal = "Frontiers in Big Data",

issn = "2624-909X",

publisher = "Frontiers Media SA",

}

Deiana, AMC, Tran, N, Agar, J, Blott, M, Di Guglielmo, G, Duarte, J, Harris, P, Hauck, S, Liu, M, Neubauer, MS, Ngadiuba, J, Ogrenci-Memik, S, Pierini, M, Aarrestad, T, Bähr, S, Becker, J, Berthold, AS, Bonventre, RJ, Müller Bravo, TE, Diefenthaler, M, Dong, Z, Fritzsche, N, Gholami, A, Govorkova, E, Guo, D, Hazelwood, KJ, Herwig, C, Khan, B, Kim, S, Klijnsma, T, Liu, Y, Lo, KH, Nguyen, T, Pezzullo, G, Rasoulinezhad, S, Rivera, RA, Scholberg, K, Selig, J, Sen, S, Strukov, D, Tang, W, Thais, S, Unger, KL, Vilalta, R, von Krosigk, B, Wang, S & Warburton, TK 2022, 'Applications and Techniques for Fast Machine Learning in Science', Frontiers in Big Data, vol. 5, 787421. https://doi.org/10.3389/fdata.2022.787421

TY - JOUR

T1 - Applications and Techniques for Fast Machine Learning in Science

AU - Deiana, Allison Mc Carn

AU - Tran, Nhan

AU - Agar, Joshua

AU - Blott, Michaela

AU - Di Guglielmo, Giuseppe

AU - Duarte, Javier

AU - Harris, Philip

AU - Hauck, Scott

AU - Liu, Mia

AU - Neubauer, Mark S.

AU - Ngadiuba, Jennifer

AU - Ogrenci-Memik, Seda

AU - Pierini, Maurizio

AU - Aarrestad, Thea

AU - Bähr, Steffen

AU - Becker, Jürgen

AU - Berthold, Anne Sophie

AU - Bonventre, Richard J.

AU - Müller Bravo, Tomás E.

AU - Diefenthaler, Markus

AU - Dong, Zhen

AU - Fritzsche, Nick

AU - Gholami, Amir

AU - Govorkova, Ekaterina

AU - Guo, Dongning

AU - Hazelwood, Kyle J.

AU - Herwig, Christian

AU - Khan, Babar

AU - Kim, Sehoon

AU - Klijnsma, Thomas

AU - Liu, Yaling

AU - Lo, Kin Ho

AU - Nguyen, Tri

AU - Pezzullo, Gianantonio

AU - Rasoulinezhad, Seyedramin

AU - Rivera, Ryan A.

AU - Scholberg, Kate

AU - Selig, Justin

AU - Sen, Sougata

AU - Strukov, Dmitri

AU - Tang, William

AU - Thais, Savannah

AU - Unger, Kai Lukas

AU - Vilalta, Ricardo

AU - von Krosigk, Belina

AU - Wang, Shen

AU - Warburton, Thomas K.

N1 - Publisher Copyright: Copyright © 2022 Deiana, Tran, Agar, Blott, Di Guglielmo, Duarte, Harris, Hauck, Liu, Neubauer, Ngadiuba, Ogrenci-Memik, Pierini, Aarrestad, Bähr, Becker, Berthold, Bonventre, Müller Bravo, Diefenthaler, Dong, Fritzsche, Gholami, Govorkova, Guo, Hazelwood, Herwig, Khan, Kim, Klijnsma, Liu, Lo, Nguyen, Pezzullo, Rasoulinezhad, Rivera, Scholberg, Selig, Sen, Strukov, Tang, Thais, Unger, Vilalta, von Krosigk, Wang and Warburton.

PY - 2022/4/12

Y1 - 2022/4/12

N2 - In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

AB - In this community review report, we discuss applications and techniques for fast machine learning (ML) in science—the concept of integrating powerful ML methods into the real-time experimental data processing loop to accelerate scientific discovery. The material for the report builds on two workshops held by the Fast ML for Science community and covers three main areas: applications for fast ML across a number of scientific domains; techniques for training and implementing performant and resource-efficient ML algorithms; and computing architectures, platforms, and technologies for deploying these algorithms. We also present overlapping challenges across the multiple scientific domains where common solutions can be found. This community report is intended to give plenty of examples and inspiration for scientific discovery through integrated and accelerated ML solutions. This is followed by a high-level overview and organization of technical advances, including an abundance of pointers to source material, which can enable these breakthroughs.

KW - big data

KW - codesign

KW - coprocessors

KW - fast machine learning

KW - heterogeneous computing

KW - machine learning for science

KW - particle physics

UR - http://www.scopus.com/inward/record.url?scp=85128972935&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85128972935&partnerID=8YFLogxK

U2 - 10.3389/fdata.2022.787421

DO - 10.3389/fdata.2022.787421

M3 - Review article

C2 - 35496379

AN - SCOPUS:85128972935

SN - 2624-909X

VL - 5

JO - Frontiers in Big Data

JF - Frontiers in Big Data

M1 - 787421

ER -

Applications and Techniques for Fast Machine Learning in Science

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this