IP/ESC’09

Last week I attended a conference in Grenoble, France. For some years there has been an annual event focussed on intellectual property [IP] and the broad area hardware design reusability. This year it was extended to encompass embedded systems as well – the last of the three days was largely dedicated to this topic. I submitted some proposals for papers and two were accepted. Hence my attendance. A colleague of mine was also due to attend in order to present a keynote and another technical paper. However, he was required to travel to the US last week, so I took his place. I figured that I would have a busy day on Thursday. The event was interesting and made me think about what conferences are all about and what people expect from them …

The attendance to the embedded day did seem rather thin to me, though I do not know what the expectations were. Since it was the first year with this new focus, I guess this is to be expected. The keynote, that I presented, seemed to go well enough with a reasonable audience. But it was the presentations of the technical papers that I was more concerned about.

There were a number of parallel technical tracks, with about three 90 minute sessions in each one, which ran simultaneously with invited speakers and panel sessions. The technical sessions had a moderator to introduce the speakers and accommodated four or five presentations, which gave each presenter only 15-20 minutes, which is very limiting. There are two conflicting issues/perspectives:

• This event is following the format of a “real” academic conference, where each author submits a detailed written paper in advance, which it is expected that the audience will have studied. The presentation slot is intended for a quick summary of the paper and further clarification by means of questions and answers.
• The presenters were mostly drawn from industry – companies which, at the end of the day, want to sell something. The speakers would all want to get the most “air time” to make attendees aware of what they have to offer.

These two angles are essentially incompatible. My view is that some careful thought needs to be given to the objectives and procedures of such conferences and this information clearly communicated and consistently applied. At this event, like so many others I have attended, too many of the presentations are product pitches, with very little attempt to hide the fact.

For the record, the three technical papers that I presented covered the following:

1. Dynamic memory usage in C/C++. This was a straight tutorial. Even though we sell tools and operating systems, there were no direct references.
2. An introduction to Android for embedded applications. This was largely an introduction to what Android is all about, but I did discuss the Mentor vision of how it might progress and what we have to offer.
3. Power management in embedded systems. This was the nearest I did to a product pitch, as I was discussing our research and development in this area and described our prototype technology. There is not currently a product to sell, so I feel that this is in the spirit of a real conference.