3rd COMCHA lectures

23 Oct 2023, 09:00 → 31 Oct 2023, 20:30 Europe/Madrid

Sala de Grados (Facultad de Ciencias, Universidad de Oviedo)

Isidro González Caballero (Universidad de Oviedo), Santiago Folgueras (Universidad de Oviedo)

This is the third school on Computing Challenges for the HL-LHC era. It includes dedicated lectures on Artifical Intelgence and Machine Learning, and programming in new accelerator platforms such as GPUs and FPGAs. It also includes talks explaining the several strategies which are being adopted in different fields to cope with the computing challenges in the next decade.

The School has a fee of 150 € to be paid by bank transfer to the account indicated in the registration mail. The number of participants is limited.

 

Scientific Committee

Luca Fiorini
Santiago Folgueras Gómez
Pablo Martínez Ruiz del Árbol
Arantza Oyanguren Campos
Xavier Vilasís Cardona

Local Organising Committee

Bárbara Álvarez González
Isidro González Caballero 
Santiago Folgueras Gómez
 

COMCHAcartel_v3.png

Monday, 23 October

Introduction

1 Registration

2 Inauguration and welcome

Speaker: Santiago Folgueras (Universidad de Oviedo)

231023_Introduction.pdf

3 Inaugural talk:

Speakers: Pablo Martinez Ruiz del Arbol (Instituto de Física de Cantabria), Pablo Martínez Ruíz del Árbol (IFCA-Universidad de Cantabria)

LHCComputing.pdf

4 Topical talk: Deep learning and graph-based recosntruction for the LHCb calorimeter reconstruction challenge

Speaker: Núria Valls Canudas (La Salle, Universitat Ramon Llull)

LHCb Calorimeter reconstruction challenge.pdf

17:30 coffee break

5 Building challenge

Speaker: Santiago Folgueras (Universidad de Oviedo)

The Marshmallow Challenge.pdf

Tuesday, 24 October

6 Low-power sustenaible hardware: FPGAs (intro)

Speaker: Álvaro Navarro Tobar (CIEMAT)

COMCHA curso FPGA.pdf

Software Instructions

11:00 coffe break

7 Topical talks: Revolutionizing Modern Research with ChatGPT and Generative AI

Speaker: Steven Van Vaerenbergh (Universidad de Cantabria)

slides_comcha_chatgpt.pdf

8 Low-power sustenaible hardware: FPGAs (hands-on I)

Speaker: Álvaro Navarro Tobar (CIEMAT)

1164pkg.pdf

vhdlref.pdf

13:30 lunch break

9 Low-power sustenaible hardware: FPGAs (hands-on II)

Speaker: Álvaro Navarro Tobar (CIEMAT)

17:00 coffee break

10 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

Computing Challenge.pptx

Wednesday, 25 October

11 Low-power sustenaible hardware: FPGAs (how-to)

Speaker: Álvaro Navarro Tobar (CIEMAT)

11:00 coffee break

12 Topical talks / Applications: State-of-the-art techniques to find the optimal hyperparameters for a classifier in HEP

Speaker: Miguel Fernández Gómez (Universidade de Santiago de Compostela (ES))

Hyperparameter_Tuning_COMCHA23.pdf

13 Low-power sustenaible hardware: FPGAs (hands-on III)

Speaker: Álvaro Navarro Tobar (CIEMAT)

13:30 lunch

14 Low-power sustenaible hardware: FPGAs (hands-on IV)

Speaker: Álvaro Navarro Tobar (CIEMAT)

17:00 coffee break

15 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

Thursday, 26 October

Spiking networks and neuromorphic computing

Convener: Pietro Vischia (Universidad de Oviedo)

16 From digital networks to quantum networks, passing through neuromorphic computing (I)

Speaker: Dr Pietro Vischia (Universidad de Oviedo and Instituto de Ciencias y Tecnologías Espaciales de Asturias (ICTEA))

Slides with animations

11:00 coffee break

17 From digital networks to quantum networks, passing through neuromorphic computing (II)

Speaker: Dr Pietro Vischia (Universidad de Oviedo and Instituto de Ciencias y Tecnologías Espaciales de Asturias (ICTEA))

13:30 Lunch break

18 Spiking networks and neuromorphic computing: hands-on

Speaker: Dr Pietro Vischia (Universidad de Oviedo and Instituto de Ciencias y Tecnologías Espaciales de Asturias (ICTEA))

comcha2023.ipynb

reinstall_in_colab.txt

17:00 coffee break

19 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

Friday, 27 October

20 Introduction to Graphics Processing Units (GPUs)

Speaker: Brij Kishor Jashal (IFIC-UV)

github-link

gpu01.indiacms.res.in

Slack-invitation

11:00 coffee break

21 Topical talks / applications: Computer vision applications in industry

Speaker: Xuan González (The Next Pangea)

22 Fundamentals of Accelerated Computing with GPUs: hands-on (I)

Speaker: Brij Kishor Jashal (IFIC-UV)

13:30 lunch break

23 Fundamentals of Accelerated Computing with GPUs: hands-on (II)

Speaker: Brij Kishor Jashal (IFIC-UV)

17:00 coffee break

24 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

25 Social dinner

Menu; 

Tortilla de patata
Chorizos a la sidra
Criollos a la parrilla
Huevos cocidos
Empanada
Calamares fritos
Pollo al ajillo
Patatas tres salsas
Pimientos de Padrón

 

Surtido de postres

 

Sidra de tonel, sangría, vino de la casa o cerveza de barril

 

Café

Saturday, 28 October

AI: recent developments (Chat-GPT) and ethics

26 Ethics and AI

Speaker: Daniel López Castro (CSIC)

11:30 coffee break

27 AI applications to the industry

Speaker: Xavier Vilasís (La Salle-Universitat Ramon Llull)

28 Guided tour to Oviedo's city-center

Monday, 30 October

29 Advance concepts on GPU computing

Speaker: Luca Fiorini (IFIC-Valencia)

Link to the material (password needed)

SUMMARY:

This training session is structured in two parts:

• In the first part, we will discuss and have practical examples of writing code, compiling and running custom code on a GPU. We will see some use-cases  and how to adapt and optimize the code to the GPU.
• In the second part, we will use a GPU to perform the training of a complex NN model to perform particle pattern recognition inside a video, label them, and add a tag of the identified particles in a new video.

 

PREREQUIREMENTS

For this session, you will need the following requirements:

• Access to a host with a NVIDIA GPU. E.g. the access you have to the Artemisa cluster during the duration of the school.
• NVIDIA compiler "nvcc" installed and setup. E.g. on Artemisa:
  • export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64" 
  • export PATH="$PATH:/usr/local/cuda/bin" 
  • export CUDA_HOME=/usr/local/cuda
• Capability of opening a Jupyter notebook (.ipynb). However, it is not required to run the code examples from inside the Jupyter notebook (optional).
• A google account to access and clone a Google Colab notebook.

 

 

11:00 coffee break

30 Fundamentals of Accelerated Computing with GPUs: hands-on (III)

Speaker: Luca Fiorini (IFIC-Valencia)

31 COMCHA network highlight, motivation and future work

Speaker: Arantza Oyanguren Campos (IFIC-UV)

3rd_COMCHA_School.pdf

13:45 lunch break

32 Topical talks / Applications: Forecasting geomagnetic storm disturbances and their uncertainties using deep learning

Speaker: Daniel Conde (IFIC-UV)

GIC_forecasting_COMCHA.pdf

33 Topical talks / Applications: Quantum Computing and High Energy Physics

Speaker: Miriam Lucio Martinez (Maastricht U.)

MLucio_COMCHA_2023.pdf

34 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

17:00 Coffee break

35 Computing challenge by N-World

Participants, organized by groups, will face a computing challenge facilitated by NFQ (N-World).

Tuesday, 31 October

Student session

36 Round-table

LINK TO UPLOAD SLIDE

11:00 coffee break

37 Results of computing challenge by N-world

upload your slides here

38 Farewell