You are signed out
Join Innoget to connect with David Galinec and thousands
of innovation-driven professionals and organizations
Senior manager with +15 years of experience in a Fortune 500 high- tech company. Five years Board member in a large consumer electronics retail chain. Strong knowledge of CIS markets. Ph.D. in Computer Science, Tech-savvy and Blockchain technology enthusiast.
Lille University of Science and Technology
January 1993 - January 1996
Computer Science
Worked on a deterministic resolution model of complex real-time problems on parallel computing architectures.
Costagliola D., Flahault A., Galinec D., Garnerin P., Menares J., Valleron A.-J.
A regression model for the non-epidemic level of influenza-like syndrome has been estimated from the 55,200 cases collected between October 1984 and August 1988 using the French Communicable Diseases Computer Network. The start of a major epidemic in 1988-89 was detected early. The size of the epidemic, for the entire country, was estimated at approximatively 4.3 million cases. The excess cost of sick-leave, among those of working age, was estimated at $86 million. (American Journal of Public Health, January 1991, 81:97-99)
D. Galinec; J. -L. Dekeyser; P. Marquet
Signal processing applications are bound by severe constraints: efficiency an reliability. To obey these constraints two main approaches exist: the synchronous approach and the asynchronous one. But, in practice, each style tends to be weak where the other is strong. Work presented in this paper is part of a project that aims at combining the respective advantages of these methods, in order to make programming of large signal processing applications easier with a run-time system close to the standard communicating sequential processes (CSP) model. (Proceedings of the Third IEEE/CIE International Conference on Signal Processing, Beijing, China, October 14-18, 1996)
Galinec D., Battini F., Dekeyser J.-L., Marquet P.
We present in this paper a new approach to the problem of programming complex real-time applications on distributed heterogeneous architectures. Assuming that the synchronous and asynchronous models have both significant advantages, we propose to merge them in a single framework. Indeed, the synchronous model provides the properest formal context to reason about time, when the asynchronous approach corresponds to a very attractive execution model. (Proceedings of the International Conference on Telecommunication, Distribution and Parallelism, Hermès, La Londe Les Maures, France, June 26- 28, 1996.)