We've got a few more iterations of Moore's law for sure. After that progress will likely happen in jumps and address non-xtor bottlenecks like memory access. E.g. wafer scale integration, 3D systems, photonics, etc.