Inverted CERN School of Computing 2023

6 Mar 2023, 09:00 → 9 Mar 2023, 17:30 Europe/Zurich

31/3-004 - IT Amphitheatre (CERN)

Alberto Pace (CERN), Kristina Gunne (CERN), Jarek Polok (CERN)

The 14th Inverted CERN School of Computing (iCSC 2023) consists of classes (lectures, exercises, demonstration and consultations) given by former CERN School of Computing students. The Inverted School provides a platform to share their knowledge by turning students into teachers. More information on the Inverted CSC events can be found at https://csc.web.cern.ch/schools/inverted-school/.

The school will take place on March 6-9, 2023 as a hybrid event - at CERN and on Zoom.  The event will be recorded.

Registrations are closed.

 

Topics covered this year include:

Accelerated Computing Neural Networks C++ compilers Multiplatform python programming

Authentication and Authorization Message Passing Systems Machine Learining Ops CPU Performance Programming

Cloud & Containers Track Finding Quantum Computing

 

iCSC2023 poster

Monday, 6 March

09:20 → 09:30 Introduction

Speaker: Alberto Pace (CERN)

Recording

Video preview

09:30 → 10:30 The most beautiful line you can draw with Kalman filter (1/2)

Track fitting is an everyday repetitive task in the high energy physics detector reconstruction chains. The precision and stability of the fitter depend on the available computing resources. A fit might cost up to half of the CPU time, that is spent on reconstruction. Kalman filters are a widespread solution for the track fitting. A classical Kalman filter is a powerful tool, that is applicable to the linear problems with Gaussian-like errors. However, in reality one has to deal with non-linear problems and sometimes with non-Gaussian errors. The numerical overheat results in instabilities and slows down the convergence. Physics and reparametrisation can help to improve the fit performance. Starting from the simple Kalman filter, we build up a more realistic Kalman filter, discussing practical tricks and possible issues of implementation. We then talk about implementation differences if using CPU or GPU.

Speaker: Valeriia Lukashenko (Nikhef National institute for subatomic physics (NL))

Recording

The most beautiful line you can draw with Kalman filter - Valeriia Lukashenko.pdf

Video preview

10:30 → 10:45 Coffee break

10:45 → 11:15 Opening welcome session

Speakers: Alberto Pace (CERN), Enrica Maria Porcari (CERN)

2023-03-06-iCSC-Introduction-A-Pace-v06.pdf

2023-03-06-iCSC-Introduction-A-Pace-v06.pptx

Recording

Video preview

11:15 → 12:15 Cloud & Containers - Everything you need to know

These days, the "cloud" is the default environment for deploying new applications.
Frequently cited benefits are lower cost, greater elasticity and less maintenance overhead.
However, for many people "using the cloud" means following obscure deployment steps that might seem like black magic.

This course aims to make newcomers familiar with cloud-native technology (building container images, deploying applications on Kubernetes etc.) as well as explain the fundamental concepts of the tech (microservices, separation of concerns and least privileges, fault tolerance).
In particular, the following topics of application development will be
covered:
BUILDING; writing applications in a cloud-native way (e.g. to work in an immutable environment) and creating container images according to best-practices;
DEPLOYING; using infrastructure-as-code to describe the application deployment (e.g. Helm charts) and using advanced features such as rolling updates and auto-scaling;
MONITORING; after multiple containers have been deployed, it is important to keep track of their status and the interaction between the services.

Speaker: Jack Henschel (CERN)

Cloud & Containers - Everything you need to know - Jack Henschel.pdf

Recording

Video preview

12:15 → 13:30 Lunch break

13:30 → 14:30 Everything that can go wrong in a message passing system

Message passing is a technique which allows to implement very performant processing software by splitting computation in pipelines and parallel nodes. However, with the great scalability comes the cost of complexity which might make such a system difficult to understand, develop and maintain. The lecture will cover the basic principles of message passing in data processing systems and typical problems that may occur when implementing and using such kind of software.

Speaker: Piotr Konopka (CERN)

Everything that can go wrong in a message passing system - Piotr Konopka.pdf

Recording

Video preview

14:30 → 15:30 The most beautiful line you can draw with Kalman filter (2/2)

Track fitting is an everyday repetitive task in the high energy physics detector reconstruction chains. The precision and stability of the fitter depend on the available computing resources. A fit might cost up to half of the CPU time, that is spent on reconstruction. Kalman filters are a widespread solution for the track fitting. A classical Kalman filter is a powerful tool, that is applicable to the linear problems with Gaussian-like errors. However, in reality one has to deal with non-linear problems and sometimes with non-Gaussian errors. The numerical overheat results in instabilities and slows down the convergence. Physics and reparametrisation can help to improve the fit performance. Starting from the simple Kalman filter, we build up a more realistic Kalman filter, discussing practical tricks and possible issues of implementation. We then talk about implementation differences if using CPU or GPU.

In these two lectures, we start from the points on planes and follow the entire track-fitting chain up to the high-level particle parameters. We discuss the connection between the geometry of the detector and the track model, as well as, the track-fitting chain. We also discuss physics-driven optimization of the algorithms based on the effect of the changes on the high-level parameters.
In the end, we discuss possible implementations of track fitting on CPU and GPU, highlighting the importance of a trade-off between speed and precision.

Speaker: Valeriia Lukashenko (Nikhef National institute for subatomic physics (NL))

Recording

The most beautiful line you can draw with Kalman filter - Valeriia Lukashenko.pdf

Video preview

15:30 → 15:45 Coffee break

15:45 → 16:45 Exercise: Cloud & Containers - Everything you need to know (1/2)

This exercise prerequisites are:

• CERN account subscribed to LXPLUS service - please check at: resource portal -> List Services
• Subscription to subordinate-users egroup - please check at: e-groups portal -> Members

Speaker: Jack Henschel (CERN)

Exercise: Cloud & Containers - Jack Henschel.pdf

Exercise instructions

16:45 → 17:45 Exercise: Cloud & Containers - Everything you need to know (2/2)

This exercise prerequisites are:

• CERN account subscribed to LXPLUS service - please check at: resource portal -> List Services
• Subscription to subordinate-users egroup - please check at: e-groups portal -> Members

Speaker: Jack Henschel (CERN)

Exercise instructions

Tuesday, 7 March

09:00 → 10:00 Authentication and Authorization for the WLCG

This lecture will introduce the concepts of authentication and authorisation and their importance to modern research infrastructures. This will then be built upon by providing an overview of the existing WLCG authentication and authorisation infrastructure (AAI), before taking a deeper look at the token based AAI the grid is currently transitioning towards, covering the motivations for change, the technologies underpinning the design, and key workflows.
The exercise class for this lecture will provide attendees with the opportunity to obtain tokens from an issuer, and then extract information from the token. This will build upon concepts from the lecture and give hands-on experience with the technologies underpinning the future of the WLCG AAI.

Speaker: Mr Tom Dack (Science and Technology Facilities Council STFC (GB))

Authentication and Authorisation for the WLCG - Tom Dack.pdf

Authentication and Authorisation for the WLCG - Tom Dack.pptx

Recording

Video preview

10:00 → 11:00 A Crash Course on Reinforcement Learning

Supervised and unsupervised machine learning has shown great performance in finding mappings between probability distributions, as e.g. in classification problems or for artificial data generation. A more difficult class of problems is decision-making, e.g. controlling dynamical systems or building mathematical algorithms because the framework requires additional time-ordering. Reinforcement learning (RL) was successful in solving such problems, e.g. in finding strategies for games, optimizing algorithms for high-performance computing, and controlling magnetic fields for nuclear fusion reactors and particle accelerators. In this lecture, I will provide an introduction to the framework, with pedagogical examples, mathematical details, and applications in particle physics. In detail, I will cover: 1) Markov decision processes (MDPs) as the mathematical foundation of RL; 2) Solving small MPDs with tabular methods; 3) Solving large MDPs with policy gradient methods.

Speaker: Felix Wagner (HEPHY Vienna)

A crash course on reinforcement learning - Felix Wagner.pdf

Handout.pdf

Recording

Video preview

11:00 → 11:15 Coffee break

11:15 → 12:15 Exercise: Authentication and Authorization for the WLCG

This exercise prerequisites are:

• Creation of Glitch account: Create Glitch account
• Creation of IRIS IAM account: Apply for an account (use eduGAIN or Local account, enter 'iCSC exercises' in Notes field)

Speaker: Mr Tom Dack (Science and Technology Facilities Council STFC (GB))

Exercise instructions

12:15 → 13:30 Lunch break

13:30 → 14:30 A simple introduction to accelerated computing

The use of hardware accelerators in High Energy Physics (HEP) is becoming increasingly popular since they are able to significantly reduce the computational time and CPU resources needed for processing and analyzing data. This lecture aims to familiarize the audience with the concept of hardware accelerators and parallel programming. In the first part of the lecture, the concept of accelerators, co-processors and heterogeneity will be discussed, with a focus on the Graphical Processing Unit (GPU). An overview of some of the current applications of GPUs in HEP will also be presented. The second part of the lecture will serve as an introduction to CUDA, a programming model designed for general computing on GPUs.

Domain : Parallel programming

Speaker: Charis Kleio Koraka (University of Wisconsin Madison (US))

A simple introduction to accelerated computing - Charis Kleio Koraka.pdf

Recording

Video preview

14:30 → 15:30 Graph Neural Networks: From fundamentals to Physics application

Non-Euclidean data structures are present everywhere in the physical and digital world. Over the last few years, an increasing number of scientific fields have started to leverage the information contained in such data structures with the advent of Geometric Deep Learning. This is also true for High Energy Physics, where Graph Neural Networks are nowadays developed and used for various tasks in different reconstruction steps.
In this lecture we will first demonstrate the expressive power of graphs as a data structure and introduce the fundamental concepts of graph theory. Then we will discuss Graph Neural Networks and lay the mathematical foundation of the most important neural mechanisms such as Neural Message Passing or Graph Convolution. Lastly we will examine applications of Graph Neural Networks in High Energy Physics that make use of the aforementioned technologies.

This lecture aims at the particle physicist who approaches Graph Neural Networks as a practitioner. The main objectives are to illustrate the reasons that Graph Neural Networks are powerful deep learning tools and to present the minimum knowledge needed to conduct research in the computer science literature and apply established technologies to HEP.

Speaker: Ilias Tsaklidis (University of Bonn)

Graph neural Networks_ From fundamentals to physics application - Ilias Tsaklidis.pdf.pdf

Recording

Video preview

Video preview

15:30 → 15:45 Coffee break

15:45 → 16:45 How a real-world C++ compiler works

The C++ language is widely used for state-of-the-art physics analysis code. Source code must be compiled before it can be executed, which involves a number of steps. Although compiler theory is taught in most undergraduate CS courses, real-world compilers carry an aura of mysterious, highly complex software products.

This lecture aims to uncover some of those secrets by feeding snippets of C++ code to a compiler, illustrating the different processing steps and dissecting the internal representations, from source to a final binary.

Speaker: Martin Cejp (CERN)

How a real-world C++ compiler works - Martin Cejp.pdf

Recording

Video preview

16:45 → 17:45 Exercise: A simple introduction to accelerated computing

This exercise prerequisites are:

• CERN account subscribed to LXPLUS service, please check at: resource portal -> List Services

Speaker: Charis Kleio Koraka (University of Wisconsin Madison (US))

Exercise instructions

Wednesday, 8 March

09:00 → 10:00 Multiplatform Programming with Python

In this course the students can learn how to write platform agnostic code using Python (and some C). Some knowledge (~1 year experience) of these two languages is recommended.

The lecture will focus on how Python can easily be combined with C for CPU and GPU programming, by exploiting the advantages of both languages. The goal is to introduce 3 Python libraries that are used at CERN (e.g. in modern multiparticle simulation frameworks): CFFI, CuPy and PyOpenCL. CFFI is a library for Python-C interfacing and CPU kernel execution. CuPy and PyOpenCL are libraries for kernel execution compatible with GPUs. Additionally, there will be a short review of heterogeneous programming and a comparison of the CUDA and OpenCL programming models.

In a subsequent tutorial session the students will be able to play around with these Python libraries.

Speaker: Peter Kicsiny (EPFL)

Multiplatform Programming with Python - Peter Kicsiny.pdf

Multiplatform Programming with Python - Peter Kicsiny.pptx

Recording

Video preview

10:00 → 11:00 MLOps - Going from Good to Great

MLOps - Going from Good to Great
To build a highly-performant machine learning model is not a small feat. The process requires a well-curated dataset, a suitable algorithm as well as finely tuned hyperparameters of the very algorithm. Once an ML model reaches a certain degree of maturity and is shared with a broader user base, a new set of operational challenges come to play. The growing field of MLOps addresses these challenges to ease the friction related to model distribution. In this lecture and exercise session, we will explore and practice main MLOps aspects, including but not limited to:
1. Selection and versioning of training datasets
2. Reproducibility of models and computing environments
3. Model encapsulation with HTTP API
4. Model versioning and roll-out strategies
5. Monitoring of model performance and its drift over time

Speaker: Michal Maciejewski

230208 MLOps Getting from Good to Great.pdf

230208 MLOps Getting from Good to Great.pptx

Recording

Video preview

11:00 → 11:15 Coffee break

11:15 → 12:15 Exercise: Multiplatform programming with python

This exercise prerequisites are:

• CERN account subscribed to LXPLUS service, please check at: resources portal -> List Services
• Activated CERNbox account, please visit: CERNBox
• Access to CERN SWAN service, please visit: SWAN

Speaker: Peter Kicsiny (EPFL)

Exercise: Multiplatform programming with python - Peter Kicsiny

12:15 → 13:30 Lunch break

13:30 → 14:30 Quantum Computing (1/2)

This will be a general overview of quantum computing and what’s special about it spanning two lectures (2 hours) and two practice sessions (2 hours). The mathematical and physics basis will be covered (not extensively). There will be a discussion of the prospects, with an emphasis on High Energy Physics. There will be a brush over the shortcomings of quantum computing and the common misrepresentation of facts about the status of the field. The practice sessions will involve using the Qiskit and Pennylane frameworks. The aim of this mini course is to inspire the students to learn more about the subject and cautiously hype them up to be interested in the CERN quantum technology initiative or getting involved with quantum technologies in general.

Speaker: Mr Ahmed Abdelmotteleb (University of Warwick (GB))

Ahmed Abdelmotteleb iCSC Quantum Computing Lecture 1.pdf

Ahmed Abdelmotteleb iCSC Quantum Computing Lecture 1.pptx

Recording

Video preview

14:30 → 15:30 Quantum Computing (2/2)

This will be a general overview of quantum computing and what’s special about it spanning two lectures (2 hours) and two practice sessions (2 hours). The mathematical and physics basis will be covered (not extensively). There will be a discussion of the prospects, with an emphasis on High Energy Physics. There will be a brush over the shortcomings of quantum computing and the common misrepresentation of facts about the status of the field. The practice sessions will involve using the Qiskit and Pennylane frameworks. The aim of this mini course is to inspire the students to learn more about the subject and cautiously hype them up to be interested in the CERN quantum technology initiative or getting involved with quantum technologies in general.

Speaker: Mr Ahmed Abdelmotteleb (University of Warwick (GB))

Ahmed Abdelmotteleb iCSC Quantum Computing Lecture 2.pdf

Ahmed Abdelmotteleb iCSC Quantum Computing Lecture 2.pptx

Recording

Video preview

15:30 → 15:45 Coffee break

15:45 → 17:15 Exercise: MLOps - Going from Good to Great

This exercise prerequisites are:

• software pre-installation, please check details at: Prerequisites.

Speaker: Michal Maciejewski

Exercise instructions

Thursday, 9 March

09:00 → 10:00 CPU Performance Profiling on Linux in the HEP Context

The Large Hadron Collider (LHC) at CERN has generated a vast amount of information from physics events, reaching peaks of TB of data per day. Many reports show that the current analysis models (and more generally, data processing interfaces) would not be able to efficiently accommodate the amount of data in the next few years. It is both the responsibility of the frameworks to provide efficient computing tools and the user's responsibility to optimally exploit these resources. The latter is of particular interest in this lecture.

The purpose of this talk is to familiarize students with mechanisms to efficiently profile the performance of C++ and Python applications, going through real-world HEP analysis. The core of the lecture will be the identification of hotspots via perf and techniques for mitigation of different kinds of bottlenecks.

Speaker: Ivan Kabadzhov (Albert Ludwig University of Freiburg)

CPU Performance Profiling on Linux in the HEP Context - Ivan Kabadzhov.pdf

Recording

Video preview

10:00 → 11:00 An introduction to Bayesian neural networks and uncertainty quantification in neural networks

Over the past few years, many advances in the field of Deep Learning (DL) have been achieved and nowadays modern DL models are starting to be deployed in our everyday life. However, for many safety-critical applications, as long as scientific research fields, the quantification of the uncertainty of DL model predictions plays a crucial role.
In this lecture, I will introduce the basics of Bayesian Neural Networks, how they can tackle the problem of estimating model uncertainty, and the most common techniques for generalizing this method to deep neural networks.

Speaker: Jacopo Talpini (Università degli Studi di Milano-Bicocca)

An Introduction to Bayesian Neural Network and Uncertainty Quantification in Deep Learning-Jacopo Talpini.pdf

Recording

Video preview

11:00 → 11:02 Thank you

Thank you!.pdf

Thank you!.pptx

11:02 → 11:15 Coffee break

11:15 → 12:15 Exercise: Quantum Computing (1/2)

This exercise prerequisites are:

• account creation at IBM Quantum Lab: Sign In to IBM Quantum
• account creation at Xanadu Cloud: Sign In to Xanadu Cloud

Speaker: Mr Ahmed Abdelmotteleb (University of Warwick (GB))

Exercise instructions

Grover_demo.zip

Pennylane.zip

Qiskit_questions.zip

Qiskit_solutions.zip

12:15 → 13:30 Lunch break

13:30 → 14:30 Exercise: Quantum Computing (2/2)

This exercise prerequisites are:

• account creation at IBM Quantum Lab: Sign In to IBM Quantum
• account creation at Xanadu Cloud: Sign In to Xanadu Cloud

Speaker: Mr Ahmed Abdelmotteleb (University of Warwick (GB))

Exercise instructions

Grover_demo.zip

Pennylane.zip

Qiskit_questions.zip

Qiskit_solutions.zip