cs.ARNov 19, 2026

FlashX: Scaling Robotic Experience Reuse for Embodied Agents with CBA-Based NAND Flash

Authors: Yuseon Choi, Sanghyeon Park, Eunjong Lee, Hanseo Lee, Sangjin Kim

Organizations: Gwangju Institute of Science and Technology (GIST), Republic of Korea

Abstract

Embodied agents are emerging as a promising paradigm for complex robotic tasks, leveraging Vision-Language-Action (VLA) and World Action Models (WAMs) to generate actions through Diffusion Transformer (DiT)-based iterative denoising. However, existing models often struggle to generalize to out-of-distribution (OOD) environments. We observe that retrieving successful experiences and using them to warm-start denoising enables agents to better adapt to unfamiliar situations while reducing computational overhead. This creates a new opportunity to improve task accuracy and inference efficiency simultaneously, in contrast to conventional acceleration techniques that trade accuracy for computation. However, realizing this opportunity requires scaling experience banks, introducing substantial storage and search overhead that challenges the capacity, bandwidth, and power constraints of edge systems. We propose FlashX, a CMOS Directly Bonded to Array (CBA)-based NAND flash architecture for scalable experience reuse in embodied agents. FlashX exploits CBA's high internal bandwidth for parallel pSLC key search, employing interleaved Compute-in-Flash (CIF) engines to sustain sensing throughput while retaining dense TLC latent storage. Proprioception-aware clustering further reduces search overhead and enables incremental experience acquisition without global reclustering. Evaluations across VLA and WAM workloads demonstrate up to 12.24%p and 7.43%p improvements in task success under dense and compute-efficient configurations, respectively. FlashX achieves 2.57×\times higher search throughput than the strongest prior NAND search architecture under pSLC sensing, with a further 9.7×\times search reduction through clustering. This delivers 1.29--3.95×\times end-to-end speedup and 1.12--2.34×\times energy efficiency improvement over SSD-based retrieval.

Figures & tables

Explore similar work

CardsList
  1. Agent-X: Full Pipeline Acceleration of On-device AI Agents

    May 11, 2026Jinha Chung, Byeongjun Shin, Jiin Kim +1LLM Inference EfficiencyOn-Device Language Model Inference

  2. What Can a VLA Learn from the Agent That Calls It? Task-Level Experience Distillation for Embodied Navigation

    Nov 12, 2026Jinze Tao, Yuxiao Han, Kangkang Dong +4Vision-Language NavigationVision-Language-Action Models

  3. FluxVLA Engine: A One-Stop VLA Engineering Platform for Embodied Intelligence

    Sep 15, 2026Yinhao Li, Weixin Mao, Zihan Lan +22Vision-Language-Action ModelsRobot Learning