Fault Tolerance of Multiple Logic Faults in SRAM-Based FPGA Systems
Abstract
The very hight levels of integration and submicron device sizes used in current and emerging VLSI technologies for SRAM-based FPGAs lead to higher occurrences of defects and operational faults. Thus, there is a critical need for fault tolerance and reconfiguration techniques for SRAM-based FPGAs to increase chip reliability with field reconfiguration. We first propose a technique utilizing the principle of master slave to tolerate logic or cells in SRAM-based FPGAs. We show that this architectural technique can be used to build redundancy for defect and fault tolerance with limited area and performance overhead. Our algorithm improves reliability of the SRAM-based FPGAs by performing two operations: TMR (triple modular redundancy) (in which CLBs are used to triplicate a logic function whose value is obtained at the voter output) and partitioning (in which the design is partitioned into a set of MSUs (master-slave unit) to reduce the amount of configuration memory required). In response to a component failure, a functionality equivalent MSU that does not rely on the faulty component replaces the affected MSU. Our technique can handle a large numbers of faults (we show tolerance of 16 logic faults in look-up tables LUTs belonging to the same MSU). Experimental results conducted on a subset of the ITC'99 benchmarks demonstrate a high level of reliability in term of fault tolerance with low hardware overhead compared to TMR which has a 5x- 6x area overhead and high power consumption.