M2RU: Memristive Minion Recurrent Unit for On-Chip Continual Learning at the Edge
This work addresses the problem of enabling efficient on-chip continual learning for edge devices, offering a scalable and energy-efficient solution with incremental improvements in hardware design.
The paper tackled the challenge of energy-intensive training and data movement for continual learning on edge platforms by introducing M2RU, a mixed-signal architecture that achieves 312 GOPS per watt and maintains accuracy within 5% of software baselines on sequential tasks.
Continual learning on edge platforms remains challenging because recurrent networks depend on energy-intensive training procedures and frequent data movement that are impractical for embedded deployments. This work introduces M2RU, a mixed-signal architecture that implements the minion recurrent unit for efficient temporal processing with on-chip continual learning. The architecture integrates weighted-bit streaming, which enables multi-bit digital inputs to be processed in crossbars without high-resolution conversion, and an experience replay mechanism that stabilizes learning under domain shifts. M2RU achieves 15 GOPS at 48.62 mW, corresponding to 312 GOPS per watt, and maintains accuracy within 5 percent of software baselines on sequential MNIST and CIFAR-10 tasks. Compared with a CMOS digital design, the accelerator provides 29X improvement in energy efficiency. Device-aware analysis shows an expected operational lifetime of 12.2 years under continual learning workloads. These results establish M2RU as a scalable and energy-efficient platform for real-time adaptation in edge-level temporal intelligence.