I was asked today to present my research to my colleagues in three slides and 10 minutes. One slide on the context, one slide on the specificity of my research and one slide on the tools I use (both technically and scientifically). The audience were the other scientists of my research department, so it is not really a popularization effort.
I did not manage to do it in less than 4 slides, but I'm quite happy
with the result, below. In particular, that's the first time that the
separation between the modeling effort and the simulation effort in
SimGrid comes to me very clearly. I think I'll reuse it in the future.
Oh, and there is 8 additional question slides, too 
That being said, I guess that these slides are very hard to follow
without the sound track. Sorry about it 
