Elvis Rojas

Publication Activity (10 Years)

Years Active: 2019-2024
Publications (10 Years): 11

Top Topics

Learning Models

Restricted Boltzmann Machine

Top Venues

Rev. Colomb. de Computación

J. Parallel Distributed Comput.

Publications

Elvis Rojas, Diego Pérez, Esteban Meneses
A characterization of soft-error sensitivity in data-parallel and model-parallel distributed deep learning. J. Parallel Distributed Comput. 190 (2024)
Hairol Romero-Sandí, Gabriel Núñez, Elvis Rojas
A snapshot of parallelism in distributed deep learning training. Rev. Colomb. de Computación 25 (1) (2024)
Gabriel Núñez, Hairol Romero-Sandí, Elvis Rojas, Esteban Meneses
A study of pipeline parallelism in deep neural networks. Rev. Colomb. de Computación 25 (1) (2024)
Elvis Rojas, Diego Pérez, Esteban Meneses
Exploring the Effects of Silent Data Corruption in Distributed Deep Learning Training. SBAC-PAD (2022)
Elvis Rojas, Michael Knobloch, Nour Daoud, Esteban Meneses, Bernd Mohr
Early Experiences of Noise-Sensitivity Performance Analysis of a Distributed Deep Learning Framework. CLUSTER (2022)
Elvis Rojas, Esteban Meneses, Terry Jones, Don Maxwell
Understanding failures through the lifetime of a top-level supercomputer. J. Parallel Distributed Comput. 154 (2021)
Elvis Rojas, Diego Pérez, Jon C. Calhoun, Leonardo Bautista-Gomez, Terry Jones, Esteban Meneses
Understanding Soft Error Sensitivity of Deep Learning Models and Frameworks through Checkpoint Alteration. CLUSTER (2021)
Elvis Rojas, Fabricio Quirós-Corella, Terry Jones, Esteban Meneses
Large-Scale Distributed Deep Learning: A Study of Mechanisms and Trade-Offs with PyTorch. CARLA (2021)
Elvis Rojas, Esteban Meneses, Terry Jones, Don Maxwell
Towards a Model to Estimate the Reliability of Large-Scale Hybrid Supercomputers. Euro-Par (2020)
Elvis Rojas, Albert Njoroge Kahira, Esteban Meneses, Leonardo Bautista-Gomez, Rosa M. Badia
A Study of Checkpointing in Large Scale Training of Deep Neural Networks. CoRR (2020)
Elvis Rojas, Esteban Meneses, Terry Jones, Don Maxwell
Analyzing a Five-Year Failure Record of a Leadership-Class Supercomputer. SBAC-PAD (2019)