Camille's picture coming soon!

Disclaimer
Name: Camille Coti
Email: camille|dot|coti||at|etsmtl|dot|ca
Position: Professor
Phone: Currently not relevant
Office A4466
Address: Département de génie logiciel et des technologies de l'information
1100 rue Notre Dame O
Montréal (Québec) H3C 1K3, Canada
Introduction

I am currently a professor cat 2 (~associate professor) at École de Technologie Supérieure, an engineering university of the Université du Québec system.

Prior to that, between 2010 and 2021 I was an assistant professor at University Sorbonne Paris Nord. I was, and am still a member of the AOC team of the LIPN. From Sept 2019 through Aug 2021, I was on a sabbatical (délégation CNRS) at the University of Oregon. I stayed in Oregon until the end of 2021 as a research associate, and I spent a year at Université du Québec à Montréal (UQAM) in 2022.

In 2009-2010 I worked as a post doc in the HPC group of the Iowa State University of Science and Technology. I worked for the Department of Mathematics under the supervision of Glenn Luecke. Prior to that I was a PhD student in INRIA's Grand Large team. My advisor was Franck Cappello. I was part of the Parallelism team of the Laboratoire de Recherche en Informatique of University Paris-Sud in Orsay. I was also a visitor at the Innovative Computing Laboratory, where I worked with George Bosilca, Aurélien Bouteiller and Pierre Lemarinier in the MPI group of Jack Dongarra's team. Even before, I did an internship in the Department of Mathematics of King's College London, where I worked with Peter Sollich.

My academic genealogy, obtained for the Math Genealogy project using Mathematics Genealogy Grapher. My Erdos number is 4, through multiple paths: Erdos -- Alon -- Tixeuil -- Peyronnet -- me, or Erdos -- Graham -- Rosenberg -- Cappello -- me, or Erdos -- Hell -- Bampis -- Finta -- me, or Erdos -- Janson -- Lavault -- Butelle -- me.

ORCID iD iconhttps://orcid.org/0000-0002-1224-7786

My ha-index is 77, which is considered as "needs to be improved", but I don't think I can improve it anymore.

Sunset in Ajaccio, les Iles Sanguinaires, 2012 - Les Iles Sanguinaires, Ajaccio, 2003 - Calanches de Piana, 2012

Publications

Full publication list here.

Click on the plus icons to unfold each list.

Talks:

Book chapters:

Latest articles in international journals:

Latest articles in international conferences:

Latest research reports:

Latest national audience reports:

Latest papers in teaching workshops (in French):

My bibTeX.

Co-authors: plus[More/Less]

Paris OpenSHMEM (POSH)

I am the author of the POSH implementation of the OpenSHMEM standard. For the moment, POSH works for shared memory systems only.

More details on the dedicated page or on my GitHub account.

POSH logo
Parallel.py (PAR)

I contribute to parallel.py (also called PAR) mostly on portability issues. parallel.py is a simple parallel and distributed job crusher written by François Berenger.

I have written a specific branch of PAR, called PARrfoo, which is expected to be more scalable. Small-scale benchmarks showed no difference, large-scale experiments are coming... soon.

 PAR logo
QCG-OMPI
QCG-OMPI has been released! More information in the QCG-OMPI section.

download QCG-OMPI (approx. 11 MB).

International Working Group on Cloud Computing Resiliency (IWGCR)

I am an active member of the International Working Group on Cloud Computing Resiliency (IWGCR). The International Working Group on Cloud Computing Resiliency (IWGCR) is a working group with a mission to monitor and analyze Cloud Computing resiliency. The working group is composed of I.T. Executives, academic researchers and industry representatives.

In March 2012 the IWGCR released a report presenting the downtime statistics of current cloud solutions and rankings of the reliability of the solutions that are currently available on the cloud market. New reports were published in June 2013 and in March 2014.

Classement

I developped a program that computes a player's future ranking in the French Tennis system.

The French tennis ranking system works with points: if you win agains a player whose ranking is higher than yours, you get more points than if your opponent has a lower ranking than yours. The tricky part is that the ranking that is taken into account is not the one that your opponent had the day you played against each other, which would be too easy. It is the one you opponent will have for the next season. So you need to know what your opponent's future ranking to make an estimation of your future ranking.

In practice, the French Tennis Federation makes multiple computations (30 to 35) until the computation stabilizes, with an algorithm that fixes oscillations.

As a consequence, in order to have a precise idea of their future ranking, players need to looks at their opponents' palmares, then their opponents' opponents' palamres, etc, recursively, until all the connected component of the graph has been explored.

The program I have developped performs requests on the Web server of the French Tennis Federation and makes a recursive exploration of the matches that involved a player and their opponents, to make a recursive estimation of his/her future ranking. The computation stops at a given depth, which is chosen by the user of the program. The deeper you go, the more precise and reliable the estimation will be.

This program is available (under CeCILL-C and LGPL licence) on GitHub.

Miscellaneous contributions
I am nerdier than 98% of all people. Are you a nerd? Click here to take the Nerd Test, get nerdy images and jokes, and talk on the nerd forum! Valid XHTML 1.0 Strict