RelayGR: Scaling Long-Sequence Generative Recommendation via Cross-Stage Relay-Race Inference

• URL: http://arxiv.org/abs/2601.01712v1
• Date: Mon, 05 Jan 2026 01:34:06 GMT
• Title: RelayGR: Scaling Long-Sequence Generative Recommendation via Cross-Stage Relay-Race Inference
• Authors: Jiarui Wang, Huichao Chai, Yuanhang Zhang, Zongjin Zhou, Wei Guo, Xingkun Yang, Qiang Tang, Bo Pan, Jiawei Zhu, Ke Cheng, Yuting Yan, Shulan Wang, Yingjie Zhu, Zhengfan Yuan, Jiaqi Huang, Yuhan Zhang, Xiaosong Sun, Zhinan Zhang, Hong Zhu, Yongsheng Zhang, Tiantian Dong, Zhong Xiao, Deliang Liu, Chengzhou Lu, Yuan Sun, Zhiyuan Chen, Xinming Han, Zaizhu Liu, Yaoyuan Wang, Ziyang Zhang, Yong Liu, Jinxin Xu, Yajing Sun, Zhoujun Yu, Wenting Zhou, Qidong Zhang, Zhengyong Zhang, Zhonghai Gu, Yibo Jin, Yongxiang Feng, Pengfei Zuo,
• Abstract summary: Real-time recommender systems execute cascades (retrieval, pre-processing, fine-grained ranking) under strict tail-latency SLOs.<n>We present RelayGR, a production system that enables in-HBM relay-race inference for GR.<n> RelayGR supports up to 1.5$times$ longer sequences and improves SLO-compliant throughput by up to 3.6$times$.
• Score: 46.66085102313264
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Real-time recommender systems execute multi-stage cascades (retrieval, pre-processing, fine-grained ranking) under strict tail-latency SLOs, leaving only tens of milliseconds for ranking. Generative recommendation (GR) models can improve quality by consuming long user-behavior sequences, but in production their online sequence length is tightly capped by the ranking-stage P99 budget. We observe that the majority of GR tokens encode user behaviors that are independent of the item candidates, suggesting an opportunity to pre-infer a user-behavior prefix once and reuse it during ranking rather than recomputing it on the critical path. Realizing this idea at industrial scale is non-trivial: the prefix cache must survive across multiple pipeline stages before the final ranking instance is determined, the user population implies cache footprints far beyond a single device, and indiscriminate pre-inference would overload shared resources under high QPS. We present RelayGR, a production system that enables in-HBM relay-race inference for GR. RelayGR selectively pre-infers long-term user prefixes, keeps their KV caches resident in HBM over the request lifecycle, and ensures the subsequent ranking can consume them without remote fetches. RelayGR combines three techniques: 1) a sequence-aware trigger that admits only at-risk requests under a bounded cache footprint and pre-inference load, 2) an affinity-aware router that co-locates cache production and consumption by routing both the auxiliary pre-infer signal and the ranking request to the same instance, and 3) a memory-aware expander that uses server-local DRAM to capture short-term cross-request reuse while avoiding redundant reloads. We implement RelayGR on Huawei Ascend NPUs and evaluate it with real queries. Under a fixed P99 SLO, RelayGR supports up to 1.5$\times$ longer sequences and improves SLO-compliant throughput by up to 3.6$\times$.

Related papers

• GEMs: Breaking the Long-Sequence Barrier in Generative Recommendation with a Multi-Stream Decoder [54.64137490632567]
  We propose a novel and unified framework designed to capture users' sequences from long-term history.<n>Generative Multi-streamers ( GEMs) break user sequences into three streams.<n>Extensive experiments on large-scale industrial datasets demonstrate that GEMs significantly outperforms state-the-art methods in recommendation accuracy.
  arXiv  Detail & Related papers  (2026-02-14T06:42:56Z)
• SRAS: A Lightweight Reinforcement Learning-based Document Selector for Edge-Native RAG Pipelines [0.0]
  We propose SRAS (Sparse Reward-Aware Selector), a lightweight document selector trained via reinforcement learning (RL) for edge-native RAG deployment.<n>SRAS learns a compact (0.76MB) policy using Proximal Policy Optimization (PPO), guided by a hybrid reward signal combining Relaxed F1 and BERTScore.<n>This work is the first to demonstrate that RL-based document selection can be made ultra-lightweight, latency-aware, and effective for on-device RAG pipelines.
  arXiv  Detail & Related papers  (2026-01-05T04:39:31Z)
• xGR: Efficient Generative Recommendation Serving at Scale [19.770951650969973]
  We propose xGR, a GR-oriented serving system that meets strict low-latency requirements under highconcurrency scenarios.<n>xGR unifies the processing of prefill and decode phases through staged and separated KV cache.<n>Experiments with real-world recommendation service datasets demonstrate that xGR achieves at least 3.49x throughput compared to the state-of-the-art baseline.
  arXiv  Detail & Related papers  (2025-12-12T12:59:38Z)
• DynaKV: Enabling Accurate and Efficient Long-Sequence LLM Decoding on Smartphones [10.813495376006427]
  Large language models (LLMs) are increasingly expected to support efficient and effective long-sequence decoding.<n>Due to limited DRAM capacity, long-seuqence LLM decoding on smartphones is constrained by the key-value cache ( KVCache)<n>We propose DynaKV, the first adaptive KVCache management approach that jointly addresses accuracy and efficiency for long-sequence decoding on smartphones.
  arXiv  Detail & Related papers  (2025-10-20T08:56:02Z)
• A Universal Framework for Compressing Embeddings in CTR Prediction [68.27582084015044]
  We introduce a Model-agnostic Embedding Compression (MEC) framework that compresses embedding tables by quantizing pre-trained embeddings.<n>Our approach consists of two stages: first, we apply popularity-weighted regularization to balance code distribution between high- and low-frequency features.<n> Experiments on three datasets reveal that our method reduces memory usage by over 50x while maintaining or improving recommendation performance.
  arXiv  Detail & Related papers  (2025-02-21T10:12:34Z)
• RelayAttention for Efficient Large Language Model Serving with Long System Prompts [59.50256661158862]
  This paper aims to improve the efficiency of LLM services that involve long system prompts. handling these system prompts requires heavily redundant memory accesses in existing causal attention algorithms. We propose RelayAttention, an attention algorithm that allows reading hidden states from DRAM exactly once for a batch of input tokens.
  arXiv  Detail & Related papers  (2024-02-22T18:58:28Z)
• HiRE: High Recall Approximate Top-$k$ Estimation for Efficient LLM Inference [68.59839755875252]
  HiRE comprises of two novel components: (i) a compression scheme to cheaply predict top-$k$ rows/columns with high recall, followed by full computation restricted to the predicted subset, and (ii) DA-TOP-$k$: an efficient multi-device approximate top-$k$ operator. We demonstrate that on a one billion parameter model, HiRE applied to both the softmax as well as feedforward layers, achieves almost matching pretraining and downstream accuracy, and speeds up inference latency by $1.47times$ on a single TPUv5e device.
  arXiv  Detail & Related papers  (2024-02-14T18:04:36Z)
• Efficient NLP Inference at the Edge via Elastic Pipelining [0.42970700836450487]
  WRX reconciles the latency/memory tension via two novel techniques. We build WRX and evaluate it against a range of NLP tasks, under a practical range of target latencies, and on both CPU and GPU.
  arXiv  Detail & Related papers  (2022-07-11T17:15:57Z)
• Self-Gated Memory Recurrent Network for Efficient Scalable HDR Deghosting [59.04604001936661]
  We propose a novel recurrent network-based HDR deghosting method for fusing arbitrary length dynamic sequences. We introduce a new recurrent cell architecture, namely Self-Gated Memory (SGM) cell, that outperforms the standard LSTM cell. The proposed approach achieves state-of-the-art performance compared to existing HDR deghosting methods quantitatively across three publicly available datasets.
  arXiv  Detail & Related papers  (2021-12-24T12:36:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.