GiPH: Generalizable Placement Learning for Adaptive Heterogeneous Computing
GiPH: Generalizable Placement Learning for Adaptive Heterogeneous Computing
复制标题
DOI：
10.48550/arxiv.2305.14562
复制
发表时间：
2023-05
期刊：
影响因子：
--
通讯作者：
Yi Hu;Chao Zhang;E. Andert;Harshul Singh;Aviral Shrivastava;J. Laudon;Yan-Quan Zhou;Bob Iannucci;Carlee Joe-Wong
中科院分区：
文献类型：
--
作者：
Yi Hu;Chao Zhang;E. Andert;Harshul Singh;Aviral Shrivastava;J. Laudon;Yan-Quan Zhou;Bob Iannucci;Carlee Joe-Wong
在目标设备集群中精心安排计算应用对于实现较短的应用完成时间至关重要。由于该问题具有NP难和组合性质，所以极具挑战性。近年来，受在云服务器上放置神经网络问题的启发，人们提出了基于学习的方法来学习一种可应用于未见过的应用的放置策略。然而，这些方法通常假定设备集群是固定的，但在移动或边缘计算环境中并非如此，在这些环境中，异构设备会因特定应用而进出范围。我们提出了一种名为GiPH的新学习方法，它通过以下方式学习可推广到动态设备集群的策略：1）一种新颖的图表示gpNet，它能有效地对选择良好放置所需的信息进行编码；2）一个可扩展的图神经网络（GNN），它学习gpNet信息的概要。GiPH将放置问题转化为寻找一系列放置改进的问题，学习一种选择该序列的策略，该策略可扩展到任意规模的问题。我们使用大量任务图和设备集群对GiPH进行评估，结果表明我们学习到的策略能快速为新的问题实例找到良好的放置方案。GiPH找到的放置方案可使完成时间降低多达30.5%，搜索速度比其他基于搜索的放置策略快多达3倍。
Careful placement of a computational application within a target device cluster is critical for achieving low application completion time. The problem is challenging due to its NP-hardness and combinatorial nature. In recent years, learning-based approaches have been proposed to learn a placement policy that can be applied to unseen applications, motivated by the problem of placing a neural network across cloud servers. These approaches, however, generally assume the device cluster is fixed, which is not the case in mobile or edge computing settings, where heterogeneous devices move in and out of range for a particular application. We propose a new learning approach called GiPH, which learns policies that generalize to dynamic device clusters via 1) a novel graph representation gpNet that efficiently encodes the information needed for choosing a good placement, and 2) a scalable graph neural network (GNN) that learns a summary of the gpNet information. GiPH turns the placement problem into that of finding a sequence of placement improvements, learning a policy for selecting this sequence that scales to problems of arbitrary size. We evaluate GiPH with a wide range of task graphs and device clusters and show that our learned policy rapidly find good placements for new problem instances. GiPH finds placements with up to 30.5% lower completion times, searching up to 3X faster than other search-based placement policies.