Experiences in the teaching of a new computing paradigm: DataFlow supercomputing

If a new paradigm is presented before experienced professionals, even if it already demonstrates superiority, the presentation is typically received with resistance. If a new paradigm already demonstrates superiority, and it is presented before students, it is typically received with enthusiasm, especially if tools are available that enable the paradigm to be utilized with a relative ease, and even more if a library of useful examples is at the disposal of new users. Over a period of time, these students grow into experienced professionals who accept the new paradigm, at their new industrial posts, without any resistance. With the above in mind, a series of courses was organized for universities in the USA and Europe, to teach the data flow computing paradigm, using the Maxeler approach as the baseline for practical work. The courses were organized on three different levels: (A) One-hour presentations, followed by another hour about the programming model, and a hands-on hour with tools, (B) A certified full-day short-course that students could use in their transcripts as a diploma supplement, and (C) A one-semester course for 1, 2, or 3 credits that students could use as a graduation requirement. Different presentation levels require different methods of presentation and produce different results in the dissemination domain. These results are here presented for 10 universities in the group A, 10 in the group B, and 10 in the group C. Where appropriate, the findings are backed with statistics data.