Sep 4 – 17, 2022
AGH University of Science and Technology
Europe/Warsaw timezone

Academic programme

The complete programme will offer over 50 hours of lectures and hands-on exercises. The programme is organized over three distinct tracks: Physics Computing, Software Engineering, and Data Technologies. In addition, guest lectures, student presentations and special evening talks will be organised. Finally, AGH (the hosting university) plans to offer an optional, half-day CUDA training.

(Please note that this programme may be subject to minor changes.)

  • Physics Computing

    Introduction to Physics Computing
    by Arnulf Quadt (University of Göttingen)
    2h lectures

    • foundations of particle physics
    • introduction to the Standard Model
    • event filtering
    • calibration and alignment
    • event reconstruction
    • event simulation
    • physics analysis
    • data flow and computing resources

    Data Science and Interactive Data Exploration
    by Bob Jacobsen (UC Berkeley)
    2h lectures + 2h exercises

    • introduction, data science tools
    • using data from different sources
    • non-numeric data

    Data Analysis
    by Toni Šćulac (University of Split)
    4h lectures + 3h exercises

    • introduction to data analysis
    • probability density functions and Monte Carlo methods
    • parameter estimation and confidence intervals
    • hypothesis testing and p-value

    Introduction to Machine Learning
    by Tomasz Szumlak (AGH) and Kamila Kalecinska (AGH)
    3h lectures + 3h exercises

    • what is machine learning
    • learning algorithm, loss function, optimisation
    • overfitting and underfitting
    • machine learning in HEP
  • Software Engineering

    Tools and Techniques
    by Bob Jacobsen (UC Berkeley)
    2h lectures + 3h exercises

    • introduction to software engineering
    • test frameworks, memory checkers
    • collaborating on complex software

    Software Design in the Many-Cores Era
    by Andrei Gheata (CERN) and Stephan Hageboeck (CERN)
    4h lectures + 3h exercises

    • Amdahl's and Gustafson's laws, data and task parallelism
    • parallel programming in C++, concurrency and synchronisation
    • performance and correctness - profiling and debugging multithreaded applications
    • patterns for parallel software development

    Creating Secure Software
    by Sebastian Lopienski (CERN)
    3h lectures + 3h exercises

    • introduction to computer security
    • security in different phases of software development
    • web application security
  • Data Technologies

    Data Management
    by Alberto Pace (CERN)
    5h lectures

    • data workflow, storage models and technologies
    • reliability and error correction
    • practical cryptography: hash functions, symmetric and asymmetric encryption, digital signatures
    • authentication, authorization and accounting: PKI, certificates, Kerberos, OpenID, OAuth etc.
    • distributed hash tables, block storage, data replication, caching

    Data and Storage Technologies
    by Andreas J. Peters (CERN)
    1h lecture + 3h exercises

    • storage technologies: present and future
    • data formats and access patterns
    • optimizations in IO systems
    • redundancy, cloud storage

    Data Visualization
    by Eamonn Maguire (Proton, Switzerland)
    2h lectures + 2h exercises

    • data visualization: theory and practical applications
    • multi-dimensional data visualization
  • Additional lectures

    Student lightning talks session

    Guest lecture
    Heterogeneous computing
    by Tomasz Szumlak (AGH)

    Special Evening talk
    When Internet history meets philosophy
    by Francois Fluckiger (CERN)

  • (Optional) CUDA training

    Fundamentals of Accelerated Computing with CUDA C/C++

    This optional half-day course will allow you to learn how to accelerate and optimise existing C/C++ CPU-only applications to leverage the power of GPUs using innovative and modern CUDA techniques. It is also an excellent way to start working with highly optimised professional tools like Nsight integrated development environment with a graphical profiler. To start your journey with the massively parallel world, you are going to need a basic C/C++ competency, including familiarity with variable types, loops, functions, arrays, etc.

    This course, kindly organized by AGH University of Science and Technology (the hosting university), is offered to the participants of CSC 2022 for free (the usual fee is approximately 100 USD per person with a non-profit academic background). The promotion code which unlocks the materials and computation time in NVIDIA cloud will be given to you at the beginning of the course. The materials can be accessed and run in the cloud for approximately six months after the course. It is possible to get an official Certificate of Competency (CoC) issued by the NVIDIA after completing the exam session (at the end of the course day, or at any convenient time up to six months after the CSC 2022).