NSDI 26 - OneSidedMW: Managing Disaggregated Memory Efficiently, Flexibly, and Securely with RNIC @UsenixOrg
NSDI 26 - OneSidedMW: Managing Disaggregated Memory Efficiently, Flexibly, and Securely with RNIC  @UsenixOrg
Uploaded June 2026 | Updated September 2026, 3 weeks ago
OneSidedMW: Managing Disaggregated Memory Efficiently, Flexibly, and Securely with RNIC Offloading

Zixuan Wang, Jinyu Gu, Xingda Wei, and Yubin Xia, Shanghai Jiao Tong University

RDMA-based memory disaggregation is gaining popularity in modern datacenters to improve memory efficiency. However, existing memory management approaches for the disaggregated memory (DM) architecture face a critical tradeoff: they either suffer from poor memory utilization due to coarse-grained allocation, or encounter significant challenges in terms of performance, memory node CPU overhead, security vulnerabilities, and limited flexibility.
In this paper, we present OneSidedMW, a novel system that combines two advanced RDMA features—RDMA NIC (RNIC) offloading and memory windows—to provide fine-grained and highly efficient one-sided memory management primitives for DM. Specifically, it leverages RNIC offloading to perform MW binding and unbinding operations, achieving remote memory allocation and deallocation without involving the memory node’s CPU. To demonstrate the efficiency of OneSidedMW, we evaluate it over two representative DM systems: swap-based systems and disaggregated key-value stores. OneSidedMW achieves up to 10.6× better performance in disaggregated key-value stores and up to 32.3% performance improvement in swap-based systems, compared with the state-of-the-art approaches.

View the full NSDI '26 program at usenix.org/conference/nsdi26/technical-sessions
NSDI 26 - OneSidedMW: Managing Disaggregated Memory Efficiently, Flexibly, and Securely with RNICNSDI 26 - HEDGE: Traffic Engineering with Probabilistic Link CapacitiesNSDI 26 - Feedback-guided Adaptive Testing of Distributed Systems DesignsNSDI 26 - Fractal: Fault-Tolerant Shell-Script DistributionNSDI 26 - PlanB: Efficient Software IPv6 Lookup with Linearized B+-TreeNSDI 26 - ZOC: Elastic and Cost-Efficient Virtual SmartNIC Architecture for Cloud Physical MachinesNSDI 26 - Enabling SLO-Aware 5G Multi-Access Edge Computing with SMECSREcon26 Americas - Talking to/with MachinesNSDI 26 - R-TCP: A Framework to Optimize TCP Performance Over Rate-Limiting NetworksNSDI 26 - ZipLLM: Efficient LLM Storage via Model-Aware Synergistic Data Deduplication andNSDI 26 - eXpressSFU: Toward Super-Scalable Video Conferencing with SmartNICsNSDI 26 - Predict, Prune, Play: Efficient Video Playback Optimization Under Device...
USENIX |

NSDI '26 - OneSidedMW: Managing Disaggregated Memory Efficiently, Flexibly, and Securely with RNIC

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER