J
Jesús Labarta
Researcher at Barcelona Supercomputing Center
Publications - 392
Citations - 10357
Jesús Labarta is an academic researcher from Barcelona Supercomputing Center. The author has contributed to research in topics: Programming paradigm & Scheduling (computing). The author has an hindex of 45, co-authored 389 publications receiving 9681 citations. Previous affiliations of Jesús Labarta include University of Barcelona & University of Tennessee.
Papers
More filters
Journal ArticleDOI
Studying the impact of the Full-Network embedding on multimodal pipelines
Armand Vilalta,Dario Garcia-Gasulla,Ferran Parés,Eduard Ayguadé,Eduard Ayguadé,Jesús Labarta,Jesús Labarta,E. Ulises Moya-Sánchez,Ulises Cortés,Ulises Cortés +9 more
TL;DR: This work is partially supported by the Joint Study Agreement no.
Journal ArticleDOI
Making the Best of Temporal Locality: Just-in-Time Renaming and Lazy Write-Back on the Cell/B.E
TL;DR: A technique called bypassing is introduced that allows CellSs to perform core-to-core Direct Memory Access (DMA) transfers for generic applications and two improvements are introduced: just-in-time renaming and lazy write-back.
Book ChapterDOI
Optimization of Condensed Matter Physics Application with OpenMP Tasking Model
Joel Criado,Marta Garcia-Gasulla,Jesús Labarta,Arghya Chatterjee,Oscar Hernandez,Raül Sirvent,Gonzalo Alvarez +6 more
TL;DR: The Density Matrix Renormalization Group (DMRG++) is a condensed matter physics application used to study superconductivity properties of materials and the task-based parallelization and optimization strategies of the Hamiltonian algorithm are presented.
Posted Content
On the Behavior of Convolutional Nets for Feature Extraction
Dario Garcia-Gasulla,Ferran Parés,Armand Vilalta,Jonathan Moreno,Eduard Ayguadé,Jesús Labarta,Ulises Cortés,Toyotaro Suzumura +7 more
TL;DR: In this paper, the authors measure the discriminative power of every single feature found within a deep CNN, when used for characterizing every class of 11 datasets, and propose a thresholding approach to discard most of it.
Book ChapterDOI
Folding: Detailed Analysis with Coarse Sampling
TL;DR: A mechanism called folding is presented that combines both instrumentation and sampling for trace-based performance analysis tools and takes advantage of long execution runs and low frequency sampling to finely detail the evolution of the user code with minimal overhead on the application.