NVREC: A High-Performance Recoverable Recommendation System for Non-Volatile Memory
Open Access
- 8 July 2026
- journal article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
Abstract
Current recommendation systems are designed with the assumption that computer memory is volatile. Thus, they periodically execute checkpointing operations to save model weights and embeddings to slower solid-state drives (SSDs) for failure recovery. Such designs cannot leverage the unique characteristics of non-volatile memory (NVM). This paper proposes a new recommendation model training framework called NVrec , which uses a hybrid memory system consisting of DRAM and NVM to support computing and in-memory checkpointing. NVrec proposes three recommendation-oriented techniques. (1) It partitions embeddings so that only popular vectors are placed in DRAM to hide NVM-induced latency. (2) It maintains a dual-version scheme for in-memory checkpointing: each vector in NVM alternates between a persistent version for crash recovery and a runtime version for training updates. (3) It employs fine-grained parallelism to hide the checkpointing overhead of model parameters in DRAM and GPU. Our experiments demonstrate that NVrec improves recommendation-model training throughput by 1.05 × -6.85 × and the checkpointing process by more than 5 × compared to existing models when evaluated on the synthetic and real-world datasets.Keywords
This publication has 18 references indexed in Scilit:
- Spitfire: A Three-Tier Buffer Manager for Volatile and Non-Volatile Memory2021
- MERCI: efficient embedding reduction on commodity hardware via sub-query memoization2021
- RecSSD: near data processing for solid state drive based recommendation inference2021
- Compiler aided checkpointing using crash-consistent data structures in NVMM systems2020
- AIBox2019
- Recommending what video to watch next2019
- Enforcing Crash Consistency of Evolving Network Analytics in Non-Volatile Main Memory Systems2019
- Deep Interest Network for Click-Through Rate Prediction2018
- Chain: tasks and channels for reliable intermittent programs2016
- NV-Heaps2011