PIFS-Rec: Process-In-Fabric-Switch for Large-Scale Recommendation System Inferences
Not on the map. Why?
Abstract (arXiv)
Deep Learning Recommendation Models (DLRMs) have become increasingly popular and prevalent in today's datacenters, consuming most of the AI inference cycles. The performance of DLRMs is heavily influenced by available bandwidth due to their large vector sizes in embedding tables and concurrent accesses. To achieve substantial improvements over existing solutions, novel approaches towards DLRM optimization are needed, especially, in the context of emerging interconnect technologies like CXL. This study delves into exploring CXL-enabled systems, implementing a process-in-fabric-switch (PIFS) solution to accelerate DLRMs while optimizing their memory and bandwidth scalability. We present an in-depth characterization of industry-scale DLRM workloads running on CXL-ready systems, identifying the predominant bottlenecks in existing CXL systems. We, therefore, propose PIFS-Rec, a PIFS-based scheme that implements near-data processing through downstream ports of the fabric switch. PIFS-Rec achieves a latency that is 3.89x lower than Pond, an industry-standard CXL-based system, and also outperforms BEACON, a state-of-the-art scheme, by 2.03x.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 24 of its 68 references that are in the map. First posted on arXiv 25 Sep 2024.
Builds on
24 of 68 references.
References · 24
- 2017Multimodal Machine Learning: A Survey and Taxonomycited by 1,004
- 2019Deep Learning Recommendation Model for Personalization and Recommendation Systemscited by 327
- 2019The Architectural Implications of Facebook's DNN-based Personalized Recommendationcited by 83
- 2019Compositional Embeddings Using Complementary Partitions for Memory-Efficient Recommendation Systemscited by 68
- 2021TT-Rec: Tensor Train Compression for Deep Learning Recommendation Modelscited by 58
- 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
- 2019TensorDIMM: A Practical Near-Memory Processing Architecture for Embeddings and Tensor Operations in Deep Learningcited by 42
- 2021RecSSD: Near Data Processing for Solid State Drive Based Recommendation Inferencecited by 35
- 2022RecShard: Statistical Feature-Based Memory Optimization for Industry-Scale Neural Recommendationcited by 28
- 2023Ramulator 2.0: A Modern, Modular, and Extensible DRAM Simulatorcited by 26
- 2020Centaur: A Chiplet-based, Hybrid Sparse-Dense Accelerator for Personalized Recommendationscited by 21
12 more
- 2022Merlin HugeCTR: GPU-accelerated Recommender System Training and Inferencecited by 20
- 2020Optimizing Deep Learning Recommender Systems' Training On CPU Cluster Architecturescited by 17
- 2021Accelerating SLIDE Deep Learning on Modern CPUs: Vectorization, Quantizations, Memory Optimizations, and Morecited by 15
- 2022TPP: Transparent Page Placement for CXL-Enabled Tiered-Memorycited by 10
- 2022Hercules: Heterogeneity-Aware Inference Serving for At-Scale Personalized Recommendationcited by 9
- 2023CAFE: Towards Compact, Adaptive, and Fast Embedding for Large-scale Recommendation Modelscited by 7
- 2023Demystifying CXL Memory with Genuine CXL-Ready Systems and Devicescited by 6
- 2022DisaggRec: Architecting Disaggregated Systems for Large-Scale Personalized Recommendationcited by 5
- 2020A Survey of Resource Management for Processing-in-Memory and Near-Memory Processing Architecturescited by 4
- 2021PowerTCP: Pushing the Performance Limits of Datacenter Networkscited by 3
- 2021FB: A Flexible Buffer Management Scheme for Data Center Switchescited by 2
- 2022Mars: Near-Optimal Throughput with Shallow Buffers in Reconfigurable Datacenter Networkscited by 1
Built on by
Citing papers per year: 2025: 1.
1 paper.