All projects

Research project · Principal Investigator

Deep learning computing platform

Equip UAI with a shared NVIDIA DGX A100 computing platform to support collaborative research and teaching in machine learning and deep learning.

My role
Responsible coordinator
Period
Nov 2022 – May 2026
Funding program
Concurso de Proyectos de Equipamiento Científico y Tecnológico Mediano 2022
Official project title
Plataforma de Cómputo para Deep Learning basada en NVIDIA DGX A100

Objective

Strengthen research and teaching at Universidad Adolfo Ibáñez through a shared GPU computing platform for machine learning, deep learning, and their applications across disciplines.

Approach

The proposed platform comprises two NVIDIA DGX A100 nodes connected through a high-speed InfiniBand network. Each node includes eight NVIDIA A100 GPUs connected through NVLink, supporting parallel computing and tensor operations.

The platform is intended to support established programming tools and libraries, including CUDA, OpenCL, TensorFlow, and PyTorch. The proposal also emphasizes energy efficiency and collaborative use of the infrastructure.

Contribution

The initiative is designed to expand access to advanced computing for research groups and students across UAI faculties. It includes collaboration with the Millennium Institute of Astrophysics (MAS) and the Millennium Institute for Foundational Research on Data (IMFD), connecting shared infrastructure with national and international research partnerships.