capa

Computer Arquitectures & high Performance Algorithms

Latest news:

Ricardo on PHILIPS Research
Ricardo Chaves will visit Jorge Guajardo Merchan at PHILIPS Research to work on embedded secure devicesJuly, 1 2008
New PhD Research member
Nuno Roma has achieved his PhD degree in the area of Video Transcoding Systems.May, 5 2008
New PhD Research member
Ricardo Chaves has achieved his PhD degree in the area of Secure Computing on Reconfigurable Systems.December, 18 2007
AMEP project
The motion estimation ASIP chip has just been implemented using 0.18µm CMOS 1P6M process from UMC foundry.October, 13 2007
Maestro2: top accessed paper
Our JOIN paper in 2006 is one of the top 5 accessed papersAugust, 7 2007
IEEE ARITH18
Efficient Method for Magnitude Comparison in RNS based on Two Pairs of Conjugate ModuliJuly, 3 2007
IEEE Computer
Caravela: A Novel Environment for data-flow-based distributed processingMay, 30 2007

On-Going Projects


The CAPA research area:

The Computer Architectures & high Performance Algorithms (CAPA) team was founded in 2000 as part of the Signal Processing Systems (SiPS) research group, within INESC-ID.

CAPA brings together faculty, researchers and students from computer science and electrical engineering, mainly from the Technical University of Lisbon but also from other European Universities and Institutes.

The CAPA team focuses its research mainly in new processing algorithms and in efficient architectures for high performance processing systems. CAPA team is involved on the design and development of computer and embedded systems, based on (re)configurable devices (FPGA) and specific integrated circuits for the applications (ASIC). It covers the following areas and topics:

  • efficient computational algorithms for signal processing and multimedia applications, which often present restrictions concerning real-time operation, high throughput and low latency time (e.g. object or sample based video coding, human vision system modeling, etc.);
  • computer architectures and micro-architectures, computer arithmetic, low power design, VLSI design tools, biochips and SOCs;
  • parallel programming techniques and tools, scheduling algorithms for mapping tasks on parallel systems, middleware for cluster network and grid.

Our principal goal is to perform academic research with industrial impact, with the final objective of transferring technology from academia to the industry. Presently we collaborate with european research labs and start-up companies.

In the last few years we have published more than an hundred papers in several international journals and conferences, we are involved in the scientific boards of several scientific and technical conferences and journals and act as reviewers of some of the most prestigious publications in those areas.