Name: Fast Machine Learning for Science Workshop 2023
Start: 2023-09-25T08:30:00+01:00
End: 2023-09-28T18:00:00+01:00
Location: Imperial College London

Fast Machine Learning for Science Workshop 2023

from Monday 25 September 2023 (08:30) to Thursday 28 September 2023 (18:00)

Monday 25 September 2023

08:30 Registration
Registration
08:30 - 09:30
Room: Blackett Laboratory, Lecture Theatre 1
09:30
09:30 - 10:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

09:30 Workshop Opening - Alex Tapper (Imperial College London) Sioni Paris Summers (CERN)

09:45 Fast Machine Learning at the Large Hadron Collider experiments - Thea Aarrestad (ETH Zurich (CH))
10:30 Break
Break
10:30 - 11:00
Room: Blackett Laboratory, Foyer
11:00
11:00 - 12:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

11:00 Fast Machine Learning at European XFEL - Steve Aplin (EuXFEL)

11:45 Bridging AI and biomedicine: Towards AI-driven scientific discoveries - Maria Brbic
12:30 Lunch
Lunch
12:30 - 13:30
13:30 High Energy Physics
High Energy Physics
13:30 - 15:00
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

13:30 Smart embedded DAQ systems for radiation instrumentation – Testbench and latest results - Audrey Corbeil Therrien (Université de Sherbrooke)

13:45 Scalable neural network models and terascale datasets for particle flow reconstruction - Farouk Mokhtar (Univ. of California San Diego (US))

14:00 Track reconstruction for the ATLAS Phase-II High-Level Trigger using Graph Neural Networks on FPGAs - Santosh Parajuli (Univ. Illinois at Urbana Champaign (US))

14:15 smartpixels: on-pixel featurization for single layer silicon tracking - Rachel Kovach-Fuentes (University of Chicago)

14:30 Portable Acceleration of CMS Mini-AOD Production with Coprocessors as a Service - William Patrick Mccormack (Massachusetts Inst. of Technology (US))

14:45 Accelerating Hadronic Calorimetry with Sparse Point-Voxel Convolutional Neural Networks - Jeffrey Krupa (Massachusetts Institute of Technology)
15:00 Break
Break
15:00 - 15:30
Room: Blackett Laboratory, Foyer
15:30 High Energy Physics
High Energy Physics
15:30 - 16:45
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

15:30 Convolutional Neural Networks for Real-Time Processing of ATLAS Liquid-Argon Calorimeter Signals - Anne-Sophie Berthold (Technische Universitaet Dresden (DE))

15:45 Fast ML inference in FPGAs for the Level-1 Scouting system at CMS - Thomas Owen James (CERN)

16:00 Yggdrasil Conifer: Latency and resource-aware decision trees for faster FPGA inference at the LHC - Andrew George Oliver

16:15 fwXmachina part 1: Classification with boosted decision trees on FPGA for L1 trigger - Tae Min Hong (University of Pittsburgh (US))

16:30 Realtime Anomaly Detection in the CMS Experiment Global Trigger Test Crate - Chang Sun (ETH Zurich (CH))
16:45 Break
Break
16:45 - 17:00
Room: Blackett Laboratory, Foyer
17:00 High Energy Physics - Lightning
High Energy Physics - Lightning
17:00 - 18:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

17:00 fwXmachina part 3: Anomaly detection with decision tree autoencoder on FPGA for L1 trigger - Stephen Roche (Saint Louis University)

17:05 Fast muon identification algorithm on FPGAs for the Phase II level 0 trigger of the ATLAS experiment - Graziella Russo (Sapienza Universita e INFN, Roma I (IT))

17:10 Neuromorphic Computing for On-Sensor Data Filtering on Smart-Pixels - Shruti R Kulkarni (Oak Ridge National Laboratory)

17:15 Intelligent experiments through real-time AI: Fast Data Processing and Autonomous Detector Control for sPHENIX and future EIC detectors - Micol Rigatti (Fermi National Accelerator Lab. (US))

17:20 Harnessing charged particle tracks in the Phase-2 CMS Level-1 Trigger with ultrafast Machine Learning - Christopher Edward Brown (Imperial College (GB))

17:25 B-tagging and Tau reconstruction in the Level-1 Trigger with real-time Machine Learning - Duc Minh Hoang (MIT)

17:30 Machine Learning based Data Compression on FPGA with HLS4ML - Pratik Jawahar (University of Manchester (UK - ATLAS))

17:35 Using NVIDIA Triton Server for Inference-as-a-Service at Fermilab - Claire Savard (University of Colorado Boulder (US))

17:40 Jets as sets or graphs: Fast jet classification on FPGAs for efficient triggering at the HL-LHC - Denis-Patrick Odagiu (ETH Zurich (CH))

17:45 Efficient and Robust Jet Tagging at the LHC with Knowledge Distillation - Ryan Liu (University of California, Berkeley)

17:50 Fast b-tagging at the high-level trigger of the ATLAS experiment - Stefano Franchellucci (Universite de Geneve (CH))

17:55 BDT for tau identification in the ATLAS Level-1 trigger - David Reikher (Tel Aviv University (IL))

18:00 A Convolutional Neural Network for topological fast selection algorithms in FPGAs for the HL-LHC upgrade of the CMS experiment - Maciej Mikolaj Glowacki (University of Bristol (GB))

18:05 Low Energy LArTPC Signal Detection using Anomaly Detection - Jovan Mitrevski (Fermi National Accelerator Lab. (US))

18:10 Graph Neural Networks on FPGAs with HLS4ML - Jan-Frederik Schulte (Purdue University (US))

18:15 Optimizing Sparse Neural Architectures for Low-Latency Anomaly Detection - Luke McDermott (UC San Diego & Modern Intelligence)

18:20 Accomodating Transformer in the FastML era of HEP-EX - Sitian Qian (Peking University (CN))
19:00 Reception
Reception
19:00 - 21:00
Room: Blackett Laboratory, Lecture Theatre 1
Tuesday 26 September 2023
09:00
09:00 - 10:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

09:00 Fast Machine Learning for accelerator control - Karin Rathsman (European Spallation Source)

09:45 Fast Machine Learning for laser wakefield acceleration - Matt Streeter
10:30 Break
Break
10:30 - 11:00
Room: Blackett Laboratory, Foyer
11:00
11:00 - 12:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

11:00 Fast ML for fusion simulation, optimization, and control - Jonathan Citrin

11:45 Machine Learning in Exoplanet Characterisation - Ingo Waldmann (UCL) Kai Hou Yip
12:30 Lunch
Lunch
12:30 - 13:30
13:30 Astrophysics, Accelerators
Astrophysics, Accelerators
13:30 - 15:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

13:30 Deep Spectral Networks: Enhancing Orbit Propagation and Determination in Astrodynamics - Sabin Anton (Imperial College London, Department of Aeronautics, Postgraduate Student)

13:45 Machine Learning Explorations in GRB Studies: From Classification to Extended Emission Identification - Keneth Stiven Garcia Cifuentes (Universidad Nacional Autónoma de México)

14:00 GWAK: Gravitational-Wave Anomalous Knowledge with Recurrent Autoencoders - Katya Govorkova (Massachusetts Inst. of Technology (US))

14:15 Tools and Results for Real-Time Deep Learning in Gravitational-Wave Physics - Eric Anton Moreno (Massachusetts Institute of Technology (US))

14:30 Real Time End to End Supernova Pointing - Maira Khan

14:45 Edge AI for accelerator controls: beam loss deblending - Jovan Mitrevski (Fermi National Accelerator Lab. (US))

15:00 Adaptive Machine Learning for Quench Prediction - Maira Khan

15:15 Real-Time Instability Tracking with Deep Learning on FPGAs in Magnetic Confinement Fusion Devices - Ryan Forelli (Lehigh University)
15:30 Break
Break
15:30 - 16:00
Room: Blackett Laboratory, Foyer
16:00 Materials Science, Medicine, Theory
Materials Science, Medicine, Theory
16:00 - 17:40
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

16:00 SAMBA: A Trainable Segmentation Web-App with Deep-Learning Powered Labelling - Ronan Docherty (Imperial College London)

16:15 Real-time Fitting and Materials Characterization in Band- Excitation Piezoresponse Force Microscopy - Veronica Obute (Drexel)

16:30 Real-Time Machine Learning in Materials Microscopy and Spectroscopy - Joshua Agar (Drexel)

16:45 A hybrid data-driven and data assimilation operational model for long term spatiotemporal forecasting: Global and regional PM2.5 forecasting - Fangxin Fang (Imperial College London)

17:00 Towards lightweight transformer-based models with multimodal data for low-latency surgical applications - Miguel Xochicale (University College London)

17:15 Approximating Many-Electron Wave Functions using Neural Networks - Matthew Foulkes (Imperial College London)

17:30 Crystallization Learning with Delaunay Triangulation - Guosheng Yin (Department of Mathematics Imperial College London) Guosheng Yin (Department of Mathematics Imperial College London)

17:35 Genomic Interpreter: A Hierarchical Genomic Deep Neural Network with 1D Shifted Window Transformer - Zehui Li
19:00 Public lecture: Algorithms and Flow: Lupe Fiasco’s Creative Use of LLMs - Lupe Fiasco
Public lecture: Algorithms and Flow: Lupe Fiasco’s Creative Use of LLMs
- Lupe Fiasco
19:00 - 20:00
Room: Blackett Lab. Lecture Theatre 1 Wasalu Jaco, professionally known as Lupe Fiasco, is a Chicago-born, Grammy Award-winning American rapper, record producer, entrepreneur, and community advocate. He is a luminary in the world of hip-hop, renowned for his thought-provoking lyrics, innovative storytelling, and unwavering commitment to social and political activism. Rising to fame in 2006, following the success of his debut album, Food & Liquor, Lupe has released eight acclaimed studio albums; his latest being Drill Music in Zion released in June 2022. His efforts to propagate Conscious Material garnered recognition as a Henry Crown Fellow and he is a recipient of the MLK Visiting Professorship Program at MIT for the 2022/2023 academic year. Lupe will talk about the use of LLMs in his creative endeavors, including co-writting a new rap song ‘Glass of Water’ with AI. Book your ticket here: https://www.imperial.ac.uk/events/166036/algorithms-and-flow-lupe-fiascos-creative-use-of-llms/
Wednesday 27 September 2023
09:00
09:00 - 10:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

09:00 Co-Design for Efficient & Adaptive ML - Yaman Umuroglu

09:45 In network ML: Inference at the Speed of Data - Noa Zilberman
10:30 Group Photo
Group Photo
10:30 - 10:45
Room: Blackett, Level 8 terrace

10:45 Break
Break
10:45 - 11:00
Room: Blackett Laboratory, Foyer
11:00
11:00 - 12:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

11:00 Need for Speed: How to harness the power of Large Language Models - Tobias Becker (Maxeler Technologies)

11:45 Deep Learning for Fast MR Imaging and Analysis - Chen Qin
12:30 Lunch
Lunch
12:30 - 13:30
13:30 Computer Science
Computer Science
13:30 - 15:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

13:30 Hardware-aware pruning of real-time neural networks with hls4ml Optimization API - Benjamin Ramhorst (Imperial College London)

13:45 Efficient Quantization of Deep Learning Models for Hardware Acceleration - Cheng ZHANG (Imperial College London) Jianyi Cheng (University of Cambridge)

14:00 High Granularity Quantization for Ultra-Fast ML Applications on FPGAs - Chang Sun (ETH Zurich (CH))

14:15 EventDetector: A Python Package for Time Series Event Detection - Menouar Azib (Akkodis)

14:30 FKeras: A Sensitivity Analysis Tool for Edge Neural Networks - Olivia Weng

14:45 Reconfigurable Fused and Branched CNN Accelerator - Rizwan Tariq Syed (IHP - Leibniz-Institut für innovative Mikroelektronik) Marko Andjelkovic (IHP - Leibniz-Institut für innovative Mikroelektronik) Markus Ulbricht (IHP - Leibniz-Institut für innovative Mikroelektronik) Milos Krstic (IHP - Leibniz-Institut für innovative Mikroelektronik)

15:00 PolyLUT: Learning Piecewise Polynomials for Ultra-Low Latency FPGA LUT-based Inference - Marta Andronic (Imperial College London)

15:15 Exploring medical applications of fast ML with a novel FPGA firmware framework - Freddie Renyard (University of Bristol)
15:30 Break
Break
15:30 - 16:00
Room: Blackett Laboratory, Foyer
16:00 Computer Science
Computer Science
16:00 - 17:30
Room: Blackett Laboratory, Lecture Theatre 1
Contributions

16:00 Running Converged HPC & AI Workloads on the Groq AI Inference Accelerator - Tobias Becker (Maxeler Technologies)

16:15 Optimizing for Imperfections in Analog Neural Computations on BrainScaleS-2 - Eric Kern (Heidelberg University) Hendrik Borras (Heidelberg University)

16:30 Implementing Machine Learning Methods on QICK Hardware for Qubit Readout - Javier Campos

16:45 Post-training ReLU Sparsification for Faster CNN Inference on FPGA Streaming Accelerators - Krish Agrawal (Imperial College London) Zhewen Yu (Imperial College London) Alexander Montgomerie-Corcoran (Imperial College London)

17:00 ATHEENA: A Toolflow for Hardware Early-Exit Network Automation - Benjamin Biggs

17:15 Simplifying Time-Series Recognition: Automated Feature Extraction and Modern Classification - Jan Zavadil (Department of Mathematics, FNSPE, Czech Technical University in Prague.)

17:20 Universal approximation theorem and error bounds for quantum neural networks and quantum reservoirs - Lukas Gonon (Imperial College London)

17:25 AI Upscaling with Super Resolution CNNs on FPGAs and ASICs - Ryan Forelli (Lehigh University)
17:30 Workshop Summary - Philip Coleman Harris (Massachusetts Inst. of Technology (US))
Workshop Summary
- Philip Coleman Harris (Massachusetts Inst. of Technology (US))
17:30 - 17:50
Room: Blackett Laboratory, Lecture Theatre 1
19:00 Dinner -- Millennium Gloucester Hotel London Kensington
Dinner -- Millennium Gloucester Hotel London Kensington
19:00 - 23:00
Thursday 28 September 2023
09:00 Coprocessors/SONIC Developers Meeting - Kevin Pedro (Fermi National Accelerator Lab. (US)) William Patrick Mccormack (Massachusetts Inst. of Technology (US))
Coprocessors/SONIC Developers Meeting
- Kevin Pedro (Fermi National Accelerator Lab. (US))
- William Patrick Mccormack (Massachusetts Inst. of Technology (US))
09:00 - 10:00
Room: Blackett Laboratory, Lecture Theatre 3 Zoom link: https://cern.zoom.us/j/63951739685?pwd=VTdITmdvOTc3V1hyK0xPa2t6cjhUdz09
QONNX Developers Meeting - Nhan Tran Jovan Mitrevski (Fermi National Accelerator Lab. (US)) Yaman Umuroglu
QONNX Developers Meeting
- Nhan Tran
- Jovan Mitrevski (Fermi National Accelerator Lab. (US))
- Yaman Umuroglu
09:00 - 10:00
Room: Blackett Laboratory, Lecture Theatre 2
10:00 hls4ml
hls4ml
10:00 - 12:00
Room: Blackett Laboratory, Lecture Theatre 2

12:00 Lunch
Lunch
12:00 - 13:00
13:00 Intel® FPGA AI Suite and AI Tensor Blocks: Empowering Real-time, Low-Latency, and Low-Power Deep Learning Inference with Intel FPGAs - Jahanzeb Ahmad Suleyman Demirsoy
Intel® FPGA AI Suite and AI Tensor Blocks: Empowering Real-time, Low-Latency, and Low-Power Deep Learning Inference with Intel FPGAs
- Jahanzeb Ahmad
- Suleyman Demirsoy
13:00 - 15:00
Room: Blackett Laboratory, Lecture Theatre 2 This two-part tutorial presents an update on Intel HLS flow and the Intel FPGA AI Suite. In the first part, we will have a 30-minute update on how the latest oneAPI tool flow for IP authoring works. In the second part we will present Intel FPGA AI Suite and groundbreaking AI Tensor Blocks newly integrated into Intel's latest FPGA device families for deep learning inference. These innovative FPGA components bring real-time, low-latency, and energy-efficient processing to the forefront, supported by the inherent advantages of Intel FPGAs, including I/O flexibility, dynamic reconfiguration, and long-term support. We delve into the Intel FPGA AI Suite, demonstrating its flexibility in achieving scalable performance and seamless integration with industry-leading frameworks like TensorFlow and PyTorch, facilitated by Quartus Prime Software. Moreover, we highlight the game-changing role of AI Tensor Blocks in enhancing deep learning inference performance. This tutorial offers both theoretical insights and practical experiences, equipping participants to leverage these advancements and revolutionize FPGA-based AI applications.
15:00 AI for Simulation: Transforming Traditional HPC with Graphcore IPUs - Alexander Titterton (Graphcore)
AI for Simulation: Transforming Traditional HPC with Graphcore IPUs
- Alexander Titterton (Graphcore)
15:00 - 17:00
Room: Blackett Laboratory, Lecture Theatre 2 More and more researchers working in fields such as drug discovery, weather forecasting, climate modelling and high-energy particle physics are looking towards AI-based approaches to enhance their applications, both in terms of accuracy and time-to-result. Furthermore, new approaches such as PINNs are revolutionising how neural networks can learn to emulate physical systems governed by well-defined yet complex equations. In this workshop we will take a look at some examples and explore how Graphcore’s IPU technology can accelerate such workflows on the cusp between HPC and AI.

Choose timezone