nVidia can use the portions of x86 that are no longer under patent, but they'd still lack MMX, 3DNow!, SSE and x86-64....which means they'd be releasing a 32bit processor, which wouldn't be a financially sound idea. You can't compete using outdated technology. Given nVidia's history of insulting their competition, I find it very unlikely that AMD will license x86-64 to nVidia. AMD only gave Intel a license due to fair competition laws. As far as nVidia using Transmeta's tech...not possible, as Intel bought the rights to ALL of Transmeta's patents when they bought out Transmeta. Having Transmeta's patents made the Core i7 processors possible. QPI used by Core i7, another form of AMD's HyperTransport, was patented by Transmeta....who also worked to help develop AMD's HyperTransport bus...
[citation][nom]bryce55[/nom]Couldn't they just call their architecture x86^-1, its not like people dont do this with patents all the time .[/citation]
The problem with calling an instruction set "x86^-1", is the fact that they can't use anything that's still covered by a valid patent....