View in EDS

Design and implementation of an architecture-aware hardware runtime for heterogeneous systems

Saved in:
Bibliographic Details
Title: Design and implementation of an architecture-aware hardware runtime for heterogeneous systems
Authors: De Haro Ruiz, Juan Miguel, Bosch Pons, Jaume, Jiménez González, Daniel
Publisher Information: Barcelona Supercomputing Center
Publication Year: 2020
Collection: Universitat Politècnica de Catalunya, BarcelonaTech: UPCommons - Global access to UPC knowledge
Subject Terms: Àrees temàtiques de la UPC::Informàtica::Arquitectura de computadors, High performance computing, heterogeneous systems, task-dependence analysis, High-performance computing, FPGA, task-based programming models, Càlcul intensiu (Informàtica)
Description: Parallel computing has become the norm to gain performance in multicore and heterogeneous systems. Many programming models allow to exploit this parallelism with easy to use tools. In this work we focus on task-based programming models. The parallelism is expressed with pieces of work called tasks that have data dependencies among them, and therefore have to be executed in a certain order. However, tasks that don’t depend on any other running task can be executed in parallel.
Document Type: conference object
File Description: 2 p.; application/pdf; application/rdf+xml; charset=utf-8
Language: English
Relation: https://hdl.handle.net/2117/331025
Availability: https://hdl.handle.net/2117/331025
Rights: Open Access ; Attribution-NonCommercial-NoDerivs 3.0 Spain
Accession Number: edsbas.69967AF6
Database: BASE
Description
Abstract:Parallel computing has become the norm to gain performance in multicore and heterogeneous systems. Many programming models allow to exploit this parallelism with easy to use tools. In this work we focus on task-based programming models. The parallelism is expressed with pieces of work called tasks that have data dependencies among them, and therefore have to be executed in a certain order. However, tasks that don’t depend on any other running task can be executed in parallel.