PreSto: An In-Storage Data Preprocessing System for Training Recommendation Models
Not on the map. Why?
Abstract (arXiv)
Training recommendation systems (RecSys) faces several challenges as it requires the "data preprocessing" stage to preprocess an ample amount of raw data and feed them to the GPU for training in a seamless manner. To sustain high training throughput, state-of-the-art solutions reserve a large fleet of CPU servers for preprocessing which incurs substantial deployment cost and power consumption. Our characterization reveals that prior CPU-centric preprocessing is bottlenecked on feature generation and feature normalization operations as it fails to reap out the abundant inter-/intra-feature parallelism in RecSys preprocessing. PreSto is a storage-centric preprocessing system leveraging In-Storage Processing (ISP), which offloads the bottlenecked preprocessing operations to our ISP units. We show that PreSto outperforms the baseline CPU-centric system with a 9.6× speedup in end-to-end preprocessing time, 4.3× enhancement in cost-efficiency, and 11.3× improvement in energyefficiency on average for production-scale RecSys preprocessing.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 23 of its 75 references that are in the map. First posted on arXiv 11 Jun 2024.
Builds on
23 of 75 references.
References · 23
- 2019Deep Learning Recommendation Model for Personalization and Recommendation Systemscited by 327
- 2019The Architectural Implications of Facebook's DNN-based Personalized Recommendationcited by 83
- 2021Software-Hardware Co-design for Fast and Scalable Training of Deep Learning Recommendation Modelscited by 78
- 2020DeepRecSys: A System for Optimizing End-To-End At-scale Neural Recommendation Inferencecited by 54
- 2019RecNMP: Accelerating Personalized Recommendation with Near-Memory Processingcited by 49
- 2020Understanding Training Efficiency of Deep Learning Recommendation Models at Scalecited by 43
- 2019TensorDIMM: A Practical Near-Memory Processing Architecture for Embeddings and Tensor Operations in Deep Learningcited by 42
- 2021tf.data: A Machine Learning Data Processing Frameworkcited by 40
- 2020Analyzing and Mitigating Data Stalls in DNN Trainingcited by 35
- 2021RecSSD: Near Data Processing for Solid State Drive Based Recommendation Inferencecited by 35
- 2021Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Trainingcited by 28
- 2020Centaur: A Chiplet-based, Hybrid Sparse-Dense Accelerator for Personalized Recommendationscited by 21
11 more
- 2021Plumber: Diagnosing and Removing Performance Bottlenecks in Machine Learning Data Pipelinescited by 15
- 2022tf.data service: A Case for Disaggregating ML Input Data Processingcited by 15
- 2020Tensor Casting: Co-Designing Algorithm-Architecture for Personalized Recommendation Trainingcited by 13
- 2022SmartSAGE: Training Large-scale Graph Neural Networks using In-Storage Processing Architecturescited by 12
- 2020Time-based Sequence Model for Personalization and Recommendation Systemscited by 11
- 2022Training Personalized Recommendation Systems from (GPU) Scratch: Look Forward not Backwardscited by 11
- 2022Hercules: Heterogeneity-Aware Inference Serving for At-Scale Personalized Recommendationcited by 9
- 2022RecD: Deduplication for End-to-End Deep Learning Recommendation Model Training Infrastructurecited by 9
- 2023MP-Rec: Hardware-Software Co-Design to Enable Multi-Path Recommendationcited by 9
- 2021RecPipe: Co-designing Models and Hardware to Jointly Optimize Recommendation Quality and Performancecited by 8
- 2023InTune: Reinforcement Learning-based Data Pipeline Optimization for Deep Recommendation Modelscited by 3
Built on by
Citing papers per year: 2025: 2.
2 papers.