Comment by Abrownn

To get a cut of the compute-cake. See: Nvidia LHR GPUs