2019 CAD Contest: System-level FPGA Routing with Timing Division Multiplexing Technique

Abstract

The time division multiplexing technique overcomes the bandwidth limitation by allowing FPGA chips to transmit multiple signals the maximum clocking frequency. With the additional multiplexers, this technique dramatically increases system-level routing capability in the FPGA-based emulator. However, the large number of virtual wires in the chip interconnection may impact emulation performance. The system-level FPGA routing tends to connect all virtual wires (signals) and considers emulation performance. At the same time, the challenge for system-level FPGA routing using time division multiplexing lies in the emulation performance.