Deep-Learning-Driven Prefetching for Far Memory

• URL: http://arxiv.org/abs/2506.00384v1
• Date: Sat, 31 May 2025 04:27:22 GMT
• Title: Deep-Learning-Driven Prefetching for Far Memory
• Authors: Yutong Huang, Zhiyuan Guo, Yiying Zhang,
• Abstract summary: We present FarSight, a Linux-based far-memory system that leverages deep learning (DL) to efficiently perform accurate data prefetching.<n>Our evaluation of FarSight on four data-intensive workloads shows that it outperforms the state-of-the-art far-memory system by up to 3.6 times.
• Score: 4.128884162772407
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Modern software systems face increasing runtime performance demands, particularly in emerging architectures like far memory, where local-memory misses incur significant latency. While machine learning (ML) has proven effective in offline systems optimization, its application to high-frequency, runtime-level problems remains limited due to strict performance, generalization, and integration constraints. We present FarSight, a Linux-based far-memory system that leverages deep learning (DL) to efficiently perform accurate data prefetching. FarSight separates application semantics from runtime memory layout, allowing offline-trained DL models to predict access patterns using a compact vocabulary of ordinal possibilities, resolved at runtime through lightweight mapping structures. By combining asynchronous inference, lookahead prediction, and a cache-resident DL model, FarSight achieves high prediction accuracy with low runtime overhead. Our evaluation of FarSight on four data-intensive workloads shows that it outperforms the state-of-the-art far-memory system by up to 3.6 times. Overall, this work demonstrates the feasibility and advantages of applying modern ML techniques to complex, performance-critical software runtime problems.

Related papers

• 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)
• MARLIN: Mixed-Precision Auto-Regressive Parallel Inference on Large Language Models [58.3342517278868]
  This paper describes the design of Mixed-precision AutoRegressive LINear kernels. It shows that batchsizes up to 16-32 can be supported with close to maximum ($4times$) quantization speedup. MarLIN accomplishes this via a combination of techniques, such as asynchronous memory access, complex task scheduling and pipelining.
  arXiv  Detail & Related papers  (2024-08-21T16:10:41Z)
• Coarse Correspondences Boost Spatial-Temporal Reasoning in Multimodal Language Model [51.83436609094658]
  We introduce Coarse Correspondences, a simple lightweight method that enhances MLLMs' spatial-temporal reasoning with 2D images as input. Our method uses a lightweight tracking model to identify primary object correspondences between frames in a video or across different image viewpoints. We demonstrate that this simple training-free approach brings substantial gains to GPT4-V/O consistently on four benchmarks.
  arXiv  Detail & Related papers  (2024-08-01T17:57:12Z)
• vTensor: Flexible Virtual Tensor Management for Efficient LLM Serving [53.972175896814505]
  Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests. Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests.
  arXiv  Detail & Related papers  (2024-07-22T14:37:58Z)
• S3D: A Simple and Cost-Effective Self-Speculative Decoding Scheme for Low-Memory GPUs [7.816840847892339]
  Speculative decoding (SD) has attracted a significant amount of research attention due to the substantial speedup it can achieve for LLM inference. We propose Skippy Simultaneous Speculative Decoding (or S3D), a cost-effective self-speculative SD method based on simultaneous multi-token decoding and mid-layer skipping. Our method has achieved one of the top performance-memory ratios while requiring minimal architecture changes and training data.
  arXiv  Detail & Related papers  (2024-05-30T17:54:35Z)
• JORA: JAX Tensor-Parallel LoRA Library for Retrieval Augmented Fine-Tuning [16.86356520836045]
  We introduce a novel framework for PEFT-compatible fine-tuning of Llama-2 models, leveraging distributed training. Our framework uniquely utilizes JAX's just-in-time (JIT) compilation and tensor-sharding for efficient resource management. Our experiments show more than 12x improvement in runtime compared to Hugging Face/DeepSpeed implementation with four GPU while consuming less than half the VRAM per GPU.
  arXiv  Detail & Related papers  (2024-03-17T23:02:04Z)
• Online Adaptation of Language Models with a Memory of Amortized Contexts [82.02369596879817]
  Memory of Amortized Contexts (MAC) is an efficient and effective online adaptation framework for large language models. We show how MAC can be combined with and improve the performance of popular alternatives such as retrieval augmented generations.
  arXiv  Detail & Related papers  (2024-03-07T08:34:57Z)
• Optimizing Memory Mapping Using Deep Reinforcement Learning [29.48627805378257]
  This paper focuses on the memory mapping problem that occurs during compilation of machine learning programs. We introduce an approach for solving the memory mapping problem using Reinforcement Learning. We also introduce a Reinforcement Learning agent, mallocMuZero, and show that it is capable of playing this game to discover new and improved memory mapping solutions.
  arXiv  Detail & Related papers  (2023-05-11T11:55:16Z)
• NumS: Scalable Array Programming for the Cloud [82.827921577004]
  We present NumS, an array programming library which optimize NumPy-like expressions on task-based distributed systems. This is achieved through a novel scheduler called Load Simulated Hierarchical Scheduling (LSHS) We show that LSHS enhances performance on Ray by decreasing network load by a factor of 2x, requiring 4x less memory, and reducing execution time by 10x on the logistic regression problem.
  arXiv  Detail & Related papers  (2022-06-28T20:13:40Z)
• KML: Using Machine Learning to Improve Storage Systems [0.2810625954925814]
  Machine learning techniques promise to learn patterns, generalize from them, and enable optimal solutions. We develop a prototype KML architecture and apply it to two problems: optimal read and read-size values. Experiments show that KML consumes little OS resources, adds negligible latency, and yet can learn patterns that can improve I/O throughput by as much as 2.3x or 15x.
  arXiv  Detail & Related papers  (2021-11-22T21:59:50Z)

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.