Special Track 12

Near-Data Computing and Edge Storage Systems:
Architectures and In-Situ Processing

Chairs

Hui Sun

Anhui University

Xiaohui Peng

Institute of Computing Technology,
Chinese Academy of Sciences

Introduction

The rapid growth of data-intensive applications has reshaped parallel and distributed computing. Traditional cloud-centric models are increasingly inadequate for latency‑sensitive and bandwidth‑constrained edge environments, where massive data streams must be processed and stored close to their origin. This track explores the convergence of near‑data processing (NDP) and edge storage systems that tackle the data movement bottleneck. NDP reduces costly data transfers by bringing computation closer to storage, while computational storage enables in‑situ analytics on storage media. Edge storage provides essential caching, tiering, and persistence for low‑latency applications, and supports computation‑in‑storage for AI inference and real‑time analytics.

In this domain, there are several critical challenges. How should data placement and replication be optimized across heterogeneous edge‑cloud continuums, considering dynamic network conditions and device heterogeneity? How can we enable efficient near‑memory/storage computation for diverse data-intensive workloads, ranging from lightweight AI models to complex stream processing? Furthermore, how can computation and storage be coordinated within a unified framework to overcome the “storage wall” that limits performance in edge AI systems?

In this track, we invite cutting‑edge advances, architectural innovations, and practical implementations that address these challenges. We solicit contributions that span from device‑level storage designs to distributed edge fabrics, and from data placement algorithms to in‑situ processing frameworks. This track aims to advance data‑centric computing framework and pave the way for next‑generation intelligent infrastructures.

Topics

Near-Data Processing (NDP) Architectures for Edge Computing
Computational Storage Devices and In-Situ Computation
Edge Storage Systems: Architectures, Caching, and Tiered Data Management
Data Placement and Storage-Aware Task Scheduling in Edge-Cloud Continuums
Energy-Efficient Co-Design of Storage and Computation for Edge Devices

Important Date

We are pleased to announce the important dates for PDCAT 2026. Please mark your calendars!
All deadlines are based on Anywhere on Earth (AoE), at midnight on the specified date.