ReCIM-HD: A 40 nm On-Device Learning Accelerator with Hyperdimensional Computing via Importance-Based Hybrid Digital-ReRAM CIM
ReCIM-HD is an on-device learning (ODL) accelerator that integrates gradient-free hyperdimensional computing (HDC) with resistive RAM (ReRAM)-based analog compute-in-memory (CIM). ReCIM-HD introduces three key techniques: (1) ultra-low-power HDC-based edge ODL without costly back-propagation operations; (2) non-volatile ReRAM-based storage for class hypervectors and CIM similarity search, enabling energy-efficient inference and power-gating during idle periods without data loss; and (3) a hybrid digital-analog CIM inference flow that protects a small accuracy-critical subset digitally to preserve accuracy. Additionally, it employs a Kronecker HD encoder (KE) to reduce encoding overhead and a customized instruction set architecture (ISA) to support flexible programmability. Fabricated in 40 nm, ReCIM-HD achieves 44.17 TOPS/W (hybrid mode) and 75.8 TOPS/W (ReRAM CIM mode), delivering up to 20.1x higher energy efficiency than the state-of-the-art accelerator.
Authors
- Tajana Rosing (ORCID: https://orcid.org/0000-0002-6954-997X)
- Haichao Yang (ORCID: https://orcid.org/0009-0008-2138-6789)
- Gopabandhu Hota
- Gert Cauwenberghs (ORCID: https://orcid.org/0000-0002-3166-5529)
- Soumil Jain (ORCID: https://orcid.org/0009-0009-9549-9724)
- Chang Eun Song (ORCID: https://orcid.org/0009-0002-4235-6243)
- Keming Fan (ORCID: https://orcid.org/0000-0002-6659-9971)
- Mingu Kang
- Weihong Xu
- Carlos H. Diaz
- Leo Liu
- Meng-Fan Chang
Publication Details
- Journal
- arXiv (Cornell University)
- Published
- 2026-11-18
- DOI
- https://doi.org/10.48550/arxiv.2611.23973
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- preprint