Centro Nacional de Supercomputación

The BSC-CNS mission:

Investigate, develop and manage technology to facilitate the advancement of science.
BSC-CNS objectives:
Supercomputing services and support to Spanish and European researchers
R&D in Computer, Life, Earth and Engineering Sciences
BSC-CNS is a consortium that includes:
Spanish Government        51%
Catalonian Government    37%
Universitat Politècnica de Catalunya (UPC)    12%
+300 people, 40 countries

Research areas:

COMPUTER SCIENCES 
To influence the way machines are built, programmed and used: programming models, performance tools, Big Data, computer  architecture, energy efficiency 

•Computer Architecture Operating System Interface

•Computer Architecture

•Performance Tools

•Programming Models

•Grid Computing

•Autonomic Systems and e-Business Platforms

•Storage systems
•Computer Architecture for Parallel Paradigms ⇒ Transactional Memory, Accelerating Synchronization, Low Power Vector Processors, Reliability and Fault Tolerant Architectures, Hardware Support for Big Data, Object Oriented and Task Oriented Architectures, Minimization of Energy.
•Extreme Computing
•Accelerators for HPC
•Unconventional Computer Architecture and Networks
EARTH SCIENCES

To develop and implement state-of-the-art global and regional models for short-term air quality forecast and long-term climate applications

LIFE SCIENCES
To understand living organisms by means of theoretical and computational methods (molecular modeling, genomics, proteomics)

CASE  

To develop scientific and engineering software to efficiently exploit super-computing capabilities (biomedical, geophysics, atmospheric, energy,
social and economic simulations)

 
BSC’s Contribution – Hardware Level:
Energy-Efficient Message Passing
Message passing microarchitecture
Message passing accelerator
Fast task switch
Task passing
Operation Below Safe Vdd
Automatic HW lowering of Vdd
SW-guided (low-power annotation)
Errors
Approximate Computing
Reduced error detection / correction
Reduced precision FP  / Narrow integers
Heterogeneous Computing
Architectural level
Device level

ParaDIME (Parallel Distributed Infrastructure for Minimization of Energy)
This project and the research leading to these results has received funding from European Community's Seventh Framework Programme [FP7/2007-2013] under Grant Agreement n° 319683.

For more information you can see: http://www.bsc.es/computer-sciences/computer-architecture-parallel-parad...