SDNHPC: Solutions for new challenges in high performance computing

In this project some of the challenges in high performance architectures including, on the one hand, multicore and manycore processors, GPUs, and FPGAs, and on the other cloud computing and Big Data, are considered. The proposed solutions will be validated on applications of particular interest in different areas. Thus, the challenges are organized into two sets, both at system level and at application level: Solutions for the processing of massive computing applications and HPC and cloud solutions for Big Data processing.

Objectives

The general objectives of the project are organized in two groups, which are linked by the development of optimized solutions for HPC and Cloud arqutectures:

Processing solutions for massive computing applications.
HPC and Cloud solutions for Big Data processing.

The first objective is subdivided in the following 7 subobjectives:

Analysis, modelling and optimization of the performance in multicore and manycore systems.
Efficient processing of LiDAR data in multicore and manycore architectures.
Deep Learning based techniques for multidimensional data classification in real time.
Registration techniques for multidimensional data.
Temporal analysis of multidimensional data sets.
Optimization techniques for numerical simulation over advanced architectures.
Efficient processing of biomedical images.

The second objective is divided into the following 3 subobjectives:

Big Data solutions for computing intensive applications.
New architectures for Big Data deployment.
Development of optimization techniques for numerical simulation in Cloud environments.