It is a mobile feature : there are two GPUs in the laptop. One is a small one, poor performances, low consumption and the other one is a full-fledged GPU.
When the usage of the small one is two high, the big one is switched on and used. By default, it is switched off and nvidia does not provide a way to switch it on on linux and stated they would not support it under linux.
I was under the impression that nvidia was the most linux-friendly GPU manufacturer and bought one because of that. I should have looked more into details but I was in a kind of hurry.
It was certainly the case that NVidia was the best for Linux until recently. But Intel has been working hard on their open source drivers and ATI has been catching up fast, and recently NVidia threw up their hands and said they had no plans to support Optimus under Linux.
When the usage of the small one is two high, the big one is switched on and used. By default, it is switched off and nvidia does not provide a way to switch it on on linux and stated they would not support it under linux.
I was under the impression that nvidia was the most linux-friendly GPU manufacturer and bought one because of that. I should have looked more into details but I was in a kind of hurry.