Velocity saturation is an unfortunate effect that occurs in semiconductors. In all semiconductors, there is a point at which increasing electric field no longer increases the average carrier velocity. This means that for a given charge carrier concentration, there is a maximum possible current. In silicon, the electron saturation velocity is around 1x107cm/s and the hole saturation velocity is about 0.6x107cm/s. Other semiconductors (e.g. gallium arsenide and gallium nitride) have higher saturation velocities*, leading some researchers to speculate that such semiconductors will replace silicon in the future. Personal opinion: they are wrong.
* Gallium arsenide (GaAs) is reported to have an electron saturation velocity of 2x107cm/s. This is somewhat misleading, because GaAs shows negative differential resistance. The maximum electron velocity is indeed 2x107cm/s, but increasing the electric field beyond that point decreases the electron velocity. The velocity eventually saturates at 1x107cm/s.
A highly-respected professor and chief technology officer at semiconductor foundry TSMC maintains that velocity saturation is due to optical phonon generation, but I have never understood this explanation. From an engineering perspective, it's more important to know that it happens than why it happens.