Quantifying Composable Data Center Utilization

M. Taubenblatt, A. Tantawi
{"title":"Quantifying Composable Data Center Utilization","authors":"M. Taubenblatt, A. Tantawi","doi":"10.1109/IPDPSW55747.2022.00206","DOIUrl":null,"url":null,"abstract":"Composable data centers have been a strong topic of research interest, with potential benefits including higher resource utilization, faster and independent refresh cycles, flex-ible/optimum resource allocation and better capital and oper-ational costs. In this paper we use a discrete event simulation model to provide a quantitative assessment of the utilization ben-efits for a composable system versus conventional. We consider a simple multi-server loss queue situation as well as two more realistic scenarios: a mixed memory size demand and a mixed ratio CPU-GPU demand. In all these scenarios, composable systems can provide many tens of percent utilization advantage over more conventional hardware deployments. We also explore the impact of size of the composable resource pool on utilization.","PeriodicalId":286968,"journal":{"name":"2022 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IPDPSW55747.2022.00206","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Composable data centers have been a strong topic of research interest, with potential benefits including higher resource utilization, faster and independent refresh cycles, flex-ible/optimum resource allocation and better capital and oper-ational costs. In this paper we use a discrete event simulation model to provide a quantitative assessment of the utilization ben-efits for a composable system versus conventional. We consider a simple multi-server loss queue situation as well as two more realistic scenarios: a mixed memory size demand and a mixed ratio CPU-GPU demand. In all these scenarios, composable systems can provide many tens of percent utilization advantage over more conventional hardware deployments. We also explore the impact of size of the composable resource pool on utilization.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
量化可组合数据中心利用率
可组合数据中心一直是研究兴趣的热门话题,其潜在的好处包括更高的资源利用率、更快和独立的更新周期、灵活/最佳的资源分配以及更好的资本和运营成本。在本文中,我们使用离散事件模拟模型提供了一个定量评估的利用效益的可组合系统相对于传统。我们考虑一个简单的多服务器丢失队列情况以及两个更现实的场景:混合内存大小需求和混合CPU-GPU比例需求。在所有这些场景中,与更传统的硬件部署相比,可组合系统可以提供几十个百分点的利用率优势。我们还探讨了可组合资源池的大小对利用率的影响。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
(CGRA4HPC) 2022 Invited Speaker: Pushing the Boundaries of HPC with the Integration of AI Moving from Composable to Programmable Energy-aware neural architecture selection and hyperparameter optimization Smoothing on Dynamic Concurrency Throttling An Analysis of Mapping Polybench Kernels to HPC CGRAs
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1