AlloX: Allocation across Computing Resources for Hybrid CPU/GPU clusters
AlloX: Allocation across Computing Resources for Hybrid CPU/GPU clusters
复制标题
AlloX：混合CPU/GPU集群的计算资源分配
DOI：
10.1145/3305218.3305251
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Liu, Zhenhua
中科院分区：
文献类型：
--
作者：
Le, Tan N.;Sun, Xiao;Chowdhury, Mosharaf;Liu, Zhenhua
GPU被认为是CPU的加速器。我们称这些应用程序为GPU应用程序。一些机器学习框架，如Tensorflow，支持在CPU或GPU上运行机器学习（ML）作业。Nvidia声称Titan GPU K80 12 GB可以平均加速5- 10倍。虽然GPU在性能上具有优势，但它们非常昂贵。例如，一个GPU K80的成本约为4000美元，而一个Intel Xeon E5四核的成本为350美元。传统CPU和GPU应用的共存促使云计算运营商构建混合CPU/GPU集群。虽然传统的应用程序在CPU上执行，但GPU应用程序可以在CPU或GPU上运行。在CPU/GPU集群中，如何为CPU和GPU应用提供混合CPU/GPU集群，以及如何在CPU和GPU之间分配资源？像CPU和GPU这样的可互换资源在大型集群中并不罕见。一些不同带宽的网络I/O卡如无线、以太网、InfinityBand等也可以互换。我们开发了一个工具，以在线方式估计ML作业的性能和资源（§2）。我们实现了支持资源分配的AlloX系统，并将应用程序放置在正确的资源（CPU或GPU）上，以最大限度地利用计算资源（§3）。与默认DRF相比，拟议的AlloX策略实现了高达35%的进度改进[2]。我们构建了一个模型，可以最大限度地降低CPU/GPU数据中心的总拥有成本（§4）。
GPUs are considered as the accelerators for CPUs. We call these applications GPU applications. Some machine learning frameworks like Tensorflow support their machine learning (ML) jobs running either on CPUs or GPUs. Nvidia claims that Titan GPU K80 12GB can speed up 5-10x on average. Although GPUs offer the advantages on performance, they are very expensive. For example, a GPU K80 roughly costs $4000 while an Intel Xeon E5 Quad cores costs $350.The coexist of traditional CPU and GPU applications urges cloud computing operators to build hybrid CPU/GPU clusters. While the traditional applications are executed on CPUs, the GPU applications can run on either CPUs or GPUs. In the CPU/GPU clusters, how to provision the hybrid CPU/GPU clusters for CPU and GPU applications and how to allocate the resources across CPUs and GPUs?Interchangeable resources like CPUs and GPUs are not rare in large clusters. Some network I/O cards like wireless, ethernet, infinityband with different bandwidths can also be interchangeable.In this paper, we focus on CPU/GPU systems. We develop a tool that estimates the performance and resource for an ML job in an online manner (§2). We implement AlloX system that supports resource allocation and places applications on right resources (CPU or GPU) to maximize the use of computational resource (§3). The proposed AlloX policy achieves up to 35% progress improvement compared to default DRF [2]. We build a model that minimizes the total cost of ownership for CPU/GPU data centers (§4).
可持续数据中心和需求响应的联合容量规划和运营管理
DOI：
--
发表时间：
2016
期刊：
Energy-Efficient Computing and Networking
影响因子：
--
作者：
T. Le;Zhenhua Liu;Yuan Chen;C. Bash
通讯作者：
C. Bash