Effective Algorithm-Accelerator Co-design for AI Solutions on Edge Devices

• URL: http://arxiv.org/abs/2010.07185v2
• Date: Thu, 15 Oct 2020 13:56:51 GMT
• Title: Effective Algorithm-Accelerator Co-design for AI Solutions on Edge Devices
• Authors: Cong Hao, Yao Chen, Xiaofan Zhang, Yuhong Li, Jinjun Xiong, Wen-mei Hwu and Deming Chen
• Abstract summary: High quality AI solutions require joint optimization of AI algorithms, such as deep neural networks (DNNs) and their hardware accelerators. To improve the overall solution quality as well as to boost the design productivity, efficient algorithm and accelerator co-design methodologies are indispensable. This paper emphasizes the importance and efficacy of algorithm-accelerator co-design and calls for more research breakthroughs in this interesting and demanding area.
• Score: 42.07369847938341
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: High quality AI solutions require joint optimization of AI algorithms, such as deep neural networks (DNNs), and their hardware accelerators. To improve the overall solution quality as well as to boost the design productivity, efficient algorithm and accelerator co-design methodologies are indispensable. In this paper, we first discuss the motivations and challenges for the Algorithm/Accelerator co-design problem and then provide several effective solutions. Especially, we highlight three leading works of effective co-design methodologies: 1) the first simultaneous DNN/FPGA co-design method; 2) a bi-directional lightweight DNN and accelerator co-design method; 3) a differentiable and efficient DNN and accelerator co-search method. We demonstrate the effectiveness of the proposed co-design approaches using extensive experiments on both FPGAs and GPUs, with comparisons to existing works. This paper emphasizes the importance and efficacy of algorithm-accelerator co-design and calls for more research breakthroughs in this interesting and demanding area.

Related papers

• Faster Optimal Coalition Structure Generation via Offline Coalition Selection and Graph-Based Search [61.08720171136229]
  We present a novel algorithm, SMART, for the problem based on a hybridization of three innovative techniques. Two of these techniques are based on dynamic programming, where we show a powerful connection between the coalitions selected for evaluation and the performance of the algorithms. Our techniques bring a new way of approaching the problem and a new level of precision to the field.
  arXiv  Detail & Related papers  (2024-07-22T23:24:03Z)
• Multiobjective Vehicle Routing Optimization with Time Windows: A Hybrid Approach Using Deep Reinforcement Learning and NSGA-II [52.083337333478674]
  This paper proposes a weight-aware deep reinforcement learning (WADRL) approach designed to address the multiobjective vehicle routing problem with time windows (MOVRPTW) The Non-dominated sorting genetic algorithm-II (NSGA-II) method is then employed to optimize the outcomes produced by the WADRL.
  arXiv  Detail & Related papers  (2024-07-18T02:46:06Z)
• Synergistic Dynamical Decoupling and Circuit Design for Enhanced Algorithm Performance on Near-Term Quantum Devices [0.5261718469769447]
  Dynamical decoupling (DD) is a promising technique for mitigating errors in near-term quantum devices. We analyze how hardware features and algorithm design impact the effectiveness of DD for error mitigation. The results reveal an inverse relationship between the effectiveness of DD and the inherent performance of the algorithm.
  arXiv  Detail & Related papers  (2024-05-27T14:48:05Z)
• An Efficient Learning-based Solver Comparable to Metaheuristics for the Capacitated Arc Routing Problem [67.92544792239086]
  We introduce an NN-based solver to significantly narrow the gap with advanced metaheuristics. First, we propose direction-aware facilitating attention model (DaAM) to incorporate directionality into the embedding process. Second, we design a supervised reinforcement learning scheme that involves supervised pre-training to establish a robust initial policy.
  arXiv  Detail & Related papers  (2024-03-11T02:17:42Z)
• Hybrid ACO-CI Algorithm for Beam Design problems [0.4397520291340694]
  A novel hybrid version of the Ant colony optimization (ACO) method is developed using the sample space reduction technique of the Cohort Intelligence (CI) algorithm. The proposed work could be investigate for real world applications encompassing domains of engineering, and health care problems.
  arXiv  Detail & Related papers  (2023-03-29T04:37:14Z)
• High-Speed Resource Allocation Algorithm Using a Coherent Ising Machine for NOMA Systems [3.6406488220483326]
  A key challenge to fully utilizing the effectiveness of the NOMA technique is the optimization of the resource allocation. We propose the coherent Ising machine (CIM) based optimization method for channel allocation in NOMA systems. We show that our proposed method is superior in terms of speed and the attained optimal solutions.
  arXiv  Detail & Related papers  (2022-12-03T09:22:54Z)
• An Overview and Experimental Study of Learning-based Optimization Algorithms for Vehicle Routing Problem [49.04543375851723]
  Vehicle routing problem (VRP) is a typical discrete optimization problem. Many studies consider learning-based optimization algorithms to solve VRP. This paper reviews recent advances in this field and divides relevant approaches into end-to-end approaches and step-by-step approaches.
  arXiv  Detail & Related papers  (2021-07-15T02:13:03Z)
• Auto-NBA: Efficient and Effective Search Over the Joint Space of Networks, Bitwidths, and Accelerators [29.72502711426566]
  We propose a framework dubbed Auto-NBA to enable jointly searching for the Networks, Bitwidths, and Accelerators. Our framework efficiently localizes the optimal design within the huge joint design space for each target dataset and acceleration specification. Our Auto-NBA generates networks and accelerators consistently outperform state-of-the-art designs.
  arXiv  Detail & Related papers  (2021-06-11T18:54:29Z)
• A Two-stage Framework and Reinforcement Learning-based Optimization Algorithms for Complex Scheduling Problems [54.61091936472494]
  We develop a two-stage framework, in which reinforcement learning (RL) and traditional operations research (OR) algorithms are combined together. The scheduling problem is solved in two stages, including a finite Markov decision process (MDP) and a mixed-integer programming process, respectively. Results show that the proposed algorithms could stably and efficiently obtain satisfactory scheduling schemes for agile Earth observation satellite scheduling problems.
  arXiv  Detail & Related papers  (2021-03-10T03:16:12Z)

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.