Search papers, labs, and topics across Lattice.
This paper introduces ProWAFT, a proactive fault-tolerance framework for FPGA-based CNN accelerators that leverages partial reconfiguration to optimize the application of Triple Modular Redundancy (TMR) based on workload criticality. By modeling fault propagation and reconfiguration overhead, ProWAFT dynamically selects configurations that minimize latency, energy consumption, and reliability risk. Evaluated on a Xilinx Zynq UltraScale+ platform, ProWAFT outperforms traditional static TMR and reactive recovery methods, achieving a lower composite cost while maintaining high task success rates and throughput.
ProWAFT reduces the overhead of fault tolerance in FPGA-based CNN accelerators by dynamically applying redundancy based on workload demands, achieving better performance and reliability.
SRAM-based FPGAs provide an attractive platform for energy- and latency-constrained CNN inference at the network edge, yet transient faults can lead to silent errors that compromise reliability. Always-on redundancy (e.g., full TMR) improves correctness but incurs substantial performance and energy overhead, while reactive recovery may introduce unacceptable latency on the critical path. We propose \textbf{ProWAFT}, a proactive workload-aware fault-tolerance framework for FPGA-based CNN accelerators that uses partial reconfiguration to selectively apply TMR across reconfigurable partitions. ProWAFT quantifies workload criticality, models fault propagation and reconfiguration overhead, and selects configurations that minimize a composite objective over latency, energy, and reliability risk. Implemented on a Xilinx Zynq UltraScale+ ZCU104 platform with six reconfigurable regions and evaluated on a 500-task trace derived from ResNet-18, MobileNetV2, and EfficientNet-Lite under time-varying SEU injection, ProWAFT achieves lower composite cost than static TMR and reactive reconfiguration while maintaining high task success rate and near-baseline throughput with low online decision overhead.
