The LNCC Summer Program is an event of the National Laboratory of Scientific Computing (LNCC), a research unit of the Ministry of Science, Technology and Innovation (MCTI), located in PetrĂ³polis
More info about the courses: http://veraolncc.kinghost.net/
In folders inside this there are repositories for the Santos Dumond Supercomputer School and NVIDIA/LNCC AI for Science Bootcamp and Data Science Journey. More details on the follow sections.
Santos Dumond Supercomputer School Main site
Files and repos inside this repo (modified) and outside(original):
- MC_SD01_I - SD Intro
- MC_SD02_I-II - OpenMP Intro / Vectorial and Parallel Programming Intro
- MC_SD03_I - CUDA Intro
- MC_SD03_II - CUDA Aware intro
- MC_SD04 - MPI Programming
Files and repos inside this repo (modified) and outside(original):
Two Labs were performed: F-MINST and Climate, which were also made available on the colab at: https://colab.research.google.com/drive/1iGfkeEbDQKaU7tmDigS_kR2O6P-SmPKQ?usp=sharing Lab changes were stored in the nvidia_bootcamp/labs folder.
- In Lab F-MINST the numbers of neurons and layers were varied. Then, a 2d convolutional network was used, which improved the accuracy from 0.78 to 0.90
- Lab Climate was not fully executed. Also available at https://github.com/openhackathons-org/gpubootcamp/tree/master/hpc_ai
- The Hackaton Lab was based on the improvement of a UNET network for CFD (Computational Fluid Dynamics) prediction. It was run on NVIDIA's JupyterLab (Competition.ipynb) and the changes were stored in the nvidia/hackaton folder.
- The code is also available at https://github.com/openhackathons-org/gpubootcamp/tree/master/hpc_ai , where it is noted that it is possible to use containers as well.
Data Science Journey main site
Files and repos inside this repo (modified) and outside(original):
Some notes