A few days ago, Nvidia announced the release in Linux environment of kernel modules that run their GPUs under the GPL / MIT license. Translated for humans, this means that the software that runs the most commonly used graphics cards for machine learning can be integrated even more efficiently into Linux, and without having to pay expensive licenses. This means, in other words, that with the use of “free” licenses, research will benefit not a little and that Nvidia will further consolidate its position on a market of extreme importance and criticality.
It is true that the times when Microsoft equated Linux with cancer (and then later changed their minds) are gone. As it is true that the distinction between proprietary and “free” approaches to intellectual property is no longer as Manichaean as in the past: free software and the various forms of open source licenses are a consolidated presence on the market and in research structures. However, the case of Nvidia is interesting because it is not limited to the “simple” choice to change the intellectual property regime of a software.
Mind you, it is not that Nvidia was electrocuted on the road to Damascus suddenly converting to the philosophy of the GPL: for example, the drivers user-mode like CUDA they remain fully owners. The use of free licenses for some software components is certainly part of a broader commercial (and not ideological) strategy. Therefore, it is too early to say whether this choice will push the direct competitors and, in general, the big player of the hardware market (but also of the software) to follow the path of the progressive adoption of free or open licenses for all products. Of course, and this is what interests us, we are faced with yet another recognition of the value of an idea – that of free software – which has earned a place of honor among real innovations (i.e. those that produce long-lasting positive effects duration) of our time.
On the sidelines (and in conclusion), however, there is a question to ask. Richard Stallman has been able to build a cultural revolution even before a technological one, starting from an approach that is perfectly valid in legal terms and therefore already applicable without the need for “new laws”. But why – from a strategic point of view – did he do it and not one of the many jurists who professes (va) no “experts” in the information society?