Keywords: parallel processing, optimisation, image processing, transputers, parallelisation
Start Date: 1 February 95 / Duration: 30 months
[ participants / contact]
The general objective of this research project is to design an efficient parallel programming environment for assisting the user in writing and evaluating parallel programs, which involves:
The following approaches will be used to fulfill these objectives:
After a long delay due ot difficulties in the importation process, the TN 310 system has been installed. During the waiting for installation and training, the development group prepared and attended a number of seminars and courses at UFRJ and LNCC (National Laboratory for Scientific Computing). These activities focused on distributed processing techniques and transputers technology. Further understanding on how to run the desired applications on TN 310 system were achieved. Contacts with the research laboratory of the Brazilian electricity company, CEPEL, were reinforced in order to start having some activity on this branch of the project. Close contact with CERN allowed to better discuss the physics part of the project.
A couple of seminars concerning the project were presented at UFRJ. At artificial intelligence and computing conferences, Sennape has been mentioned as part of the group's present activity. A paper concerning the project was presented at the International IFAC Workshop on Technology Transfer and Control, held in Curitiba, Brazil in 1995.
Universidade Federal do Rio de Janeiro
CP 68504 Rio de Janeiro
21945-970, BRZ
EU Partners
European Laboratory for Particle Physics (CERN), F
Telmat
Multinode, F
Non-EU Partners
Universidade Federal do Rio de Janeiro, BRZ
CEPEL,
BRZ
Petrobras, BRZ
Mr. Luiz Pereira Caloba
Tel: +55 21 280 73 93
Fax: +55 21 290
66 26
E-mail: caloba@coe.ufrj.br
SENNAPE - ITDC-143, May 1997
please address enquiries to the ESPRIT Information Desk
html version of synopsis by Nick Cook