Skip to main content

CARLA 2021 archive

Guadalajara, MexicoOctober 6 to 8, 2021Virtual edition

CARLA 2021, Latin America High Performance Computing Conference

Program

Program by Activity

Tutorials, workshops, keynotes, panels, paper sessions and posters, grouped by activity.

All times are given in Guadalajara time. The program was subject to change without notice.

Fundamental tutorials

September 27 to October 1, 2021

Tutorial

Working with a resource manager on an HPC infrastructure

Monday September 27, 08:00 to 17:00

Tutorial

OpenMP: introduction to shared memory models

Tuesday September 28, 08:00 to 13:00

Tutorial

Introduction to distributed memory models using MPI

Tuesday September 28, 13:00 to 17:00

Tutorial

Introduction to GPGPU programming in CUDA with C and C++

Wednesday September 29, 08:00 to 11:00

Tutorial

Introduction to cluster configuration and management

Thursday September 30, 08:00 to 17:00

Tutorial

OpenACC: parallel programming of heterogeneous computing systems

Friday October 1, 08:00 to 17:00

Advanced tutorials

October 11 to 15, 2021

Tutorial

Performance analysis tools

Monday October 11 and Tuesday October 12

Tutorial

The SYCL parallel programming model

Wednesday October 13, 08:00 to 12:00

Tutorial

Introduction to AI and HPC

Wednesday October 13, 13:00 to 17:00

Tutorial

Introduction to biological signal analysis

Thursday October 14, 08:00 to 17:00

Tutorial

Quantum computing

Friday October 15, 08:00 to 17:00

Keynotes and invited speakers

October 6 to 8, 2021

Keynote

Fugaku and the road to exascale

Satoshi Matsuoka · Wednesday October 6, 09:30

Invited

High performance computing with the IBM POWER10 processor

José E. Moreira · Wednesday October 6, 10:30

Invited

HPC in patient centric fusion of highly dimensional biomedical data

Natasa Przulj · Wednesday October 6, 14:30

Keynote

Machine learning in science

Katherine Yelick · Thursday October 7, 09:30

Keynote

Flex: automatic quota management

Walfredo Cirne · Thursday October 7, 10:30

Keynote

Towards a national research platform

Frank Würthwein · Friday October 8, 09:30

Invited

Computational costs of cloud computing

Lucia Drummond · Friday October 8, 10:30

Invited

Towards an ecologically viable artificial intelligence

Nayat Sánchez Pi · Friday October 8, 14:30

Panels

October 7 and 8, 2021

Panel 1

Africa and Latin American HPC and AI cooperation

Thursday October 7, 08:00. Participants: Happy Marumo Sithole, Matthew Mtumbuka, Philippe Olivier Alexandre Navaux and Luis Eliécer Cadenas. Moderator: Bianca Amaro.

Panel 2

Public policies and regulation on network infrastructures and HPC and AI services

Friday October 8, 08:00. Participants: Arturo Robles, Alejandro Patiño, Antonio García Zeballos, Natalija Gelvanovska Garcia and Gustav Kalbe. Moderator: Salma Jalife Villalón.

Tracks

Paper sessions and posters

Track HPC

HPC session 1

Wednesday October 6, 11:30 to 13:00

  • Energy consumption studies of WRF executions with the LIMITLESS monitor
  • A methodology for evaluating the energy efficiency of Post Moore architectures
  • An efficient vectorized auction algorithm for many core and multicore architectures
  • Green energy HPC data centers to improve processing cost efficiency
  • TEPUI: high performance computing infrastructure for beamlines at LNLS and Sirius
Track HPC&AI

HPC&AI session 1

Thursday October 7, 11:30 to 13:00

  • Wind prediction using deep learning and high performance computing
  • Large scale distributed deep learning: a study of mechanisms and trade offs with PyTorch
  • Distributed artificial intelligent model training and evaluation
  • An analysis of neural architecture search and hyper parameter optimization methods
Track HPC

HPC session 2

Friday October 8, 11:30 to 13:00

  • DICE: generic data abstraction for enhancing the convergence of HPC and big data
  • Improving performance of long short term memory networks for sentiment analysis using multicore and GPU architectures
  • OCFTL: an MPI implementation independent fault tolerance library for task based applications
  • A comparative study of consensus algorithms for distributed systems
  • Accelerating smart city simulations
  • Understanding COVID 19 epidemic in Costa Rica through network based modeling
Posters

Posters HPC

Thursday October 7, 16:30. Moderators: Esteban Mocskos and Juan Manuel Ramirez Alcaraz.

  • Solving the heat transfer equation by a finite difference method using multidimensional arrays in CUDA as in standard C
  • Application of artificial intelligence techniques for the optimization of vehicular flow
  • A distributed open source big data platform for real time geolocation tracking
  • Reducing cache miss rate using thread overallocation to accelerate the MPI OpenMP based 2D Hopmoc method
  • Educational real time scheduling framework on distributed mobile environments