Focus on Scalable Graphics

Interview with Xavier Cavin, founder of Scalable Graphics, supplier of software and services in parallel rendering for 3D visualization applications.

You are a young researcher, since your thesis is as recent as 2000. What made you decide to start your own company so early?

Xavier Cavin: Part of my motivation arose during the years I was working on my thesis. Upon completing studies at the Lorrain National Polytechnic Institute [Institut National Polytechnique de Lorraine], I did my doctorate as part of LORIA in Nancy, where I worked with François Cuny to develop the VSP Technology project, a start-up created in 2001. This was a way of becoming familiar with what it takes to found a company. Upon my return to France in 2002, after a post-doc at the University of Utah, I was hired by INRIA-Nancy as a research fellow manager. As a member of the EPI Alice, I worked alongside Alain Filbois, an INRIA research engineer, on the CRVHP project (Computation, Network and High-Performance Graphics) for the Lorraine state-regional planning contract. In 2005, we were joined by Christophe Mion to develop the DViz research prototype. Having just completed his engineering contract, we looked for a way to continue our collaboration. We knew that there was nothing equivalent to our know-how and we were well informed about the industry’s needs. However, I’d like to note that I have not resigned from INRIA: I am currently on leave. One might say that I “left the nest” at a fledgling stage, but creating Scalable Graphics cannot be considered a real risk in my case.

You say that all three of you were informed about the industry’s demand. How do you in fact meet a current need?

Xavier Cavin: You have to understand that the processing power for graphic cards is not unlimited. It is growing more slowly than the needs in the graphics processing sector or that of modeling. The best way to solve this problem is thus to share the capabilities of the cards and make them work at the same time. This requires a special type of technology which, as I have already explained, did not exist before we created the DViz project with the Alice project team.

So you think you can get the attention of financial players?

Xavier Cavin: There is talk every day about software for modeling, 3-D and simulation. The sectors that use them are highly varied. This ranges from the automotive sector to the 3-D reconstitution of an archaeological site, by way of design or chemistry, etc. We are often amazed at how we got into the petroleum line, but exploration of deposits requires the processing of multiple sets of data (photos, soundings, geological and seismic measurements, etc.). Thanks to many contacts we were able to make in this field, we now have financial support from our first partner: Paradigm Geosciences. In addition, we are going to take part in two key trade events this year: the meeting of the European Association of Geoscientists and Engineers (EAGE) in Rome, and then the meeting of the Society of Exploration Geophysicists (SEG) in Las Vegas. During these times of oil shortages, we are betting that our computing solutions have a bright future ahead of them.