In this paper, we have investigated the delay optimization using repeater insertion in folded graphene nanoribbon (FGNR) interconnect systems for different nanometer technology nodes. The mean free path is modeled as a function of GNR width and Fermi energy. The RLC equivalent circuit is modeled for analyzing delay and power for global interconnects. The performance of FGNR is compared to that of multi-layer horizontal graphene nanoribbon (ML-HGNR) and multi-layer vertical graphene nanoribbon (ML-VGNR) interconnects. Our analysis shows that proposed FGNR interconnect with fully diffusive edges can outperform both the ML-HGNR and ML-VGNR interconnects as well as the traditional Cu interconnects by reducing the number of repeaters.