Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors

Emerging many-core processors feature very high memory bandwidth and computational power. For example, Intel Xeon Phi many-core processors of the Knights Corner (KNC) and Knights Landing (KNL) architectures embrace 60 to 64 x86-based CPU cores with 512-bit SIMD capabilities and high-bandwidth memori...

Full description

Saved in:
Bibliographic Details
Main Authors: Cheng, Xuntao, He, Bingsheng, Lu, Mian, Lau, Chiew Tong
Other Authors: Interdisciplinary Graduate School (IGS)
Format: Article
Language:English
Published: 2020
Subjects:
Online Access:https://hdl.handle.net/10356/141470
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-141470
record_format dspace
spelling sg-ntu-dr.10356-1414702020-06-08T09:28:34Z Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors Cheng, Xuntao He, Bingsheng Lu, Mian Lau, Chiew Tong Interdisciplinary Graduate School (IGS) College of Professional and Continuing Education Engineering::Computer science and engineering In-memory Query Processing Many-core Processor Emerging many-core processors feature very high memory bandwidth and computational power. For example, Intel Xeon Phi many-core processors of the Knights Corner (KNC) and Knights Landing (KNL) architectures embrace 60 to 64 x86-based CPU cores with 512-bit SIMD capabilities and high-bandwidth memories like the GDDR5 on KNC and on-package DRAMs on KNL. In this paper, we study the performance main-memory database operators and online analytical processing (OLAP) on such many-core architectures. We find that even the state-of-the-art database operators suffer severely from memory stalls and resource underutilization on those many-core processors. We argue that a software approach decomposing a coarse-grained operator into fine-grained phases and executing two independent phases with complementary resource requirements concurrently can address this problem. This approach allows more fine-grained control of resource utilization. Our experiments demonstrate significant performance gain and high resource utilization achieved by our proposed approaches on both KNC and KNL. NRF (Natl Research Foundation, S’pore) MOE (Min. of Education, S’pore) 2020-06-08T09:28:34Z 2020-06-08T09:28:34Z 2017 Journal Article Cheng, X., He, B., Lu, M., & Lau, C. T. (2018). Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors. Journal of Parallel and Distributed Computing, 120, 395-404. doi:10.1016/j.jpdc.2017.09.005 0743-7315 https://hdl.handle.net/10356/141470 10.1016/j.jpdc.2017.09.005 2-s2.0-85039726997 120 395 404 en Journal of Parallel and Distributed Computing © 2017 Elsevier Inc. All rights reserved.
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic Engineering::Computer science and engineering
In-memory Query Processing
Many-core Processor
spellingShingle Engineering::Computer science and engineering
In-memory Query Processing
Many-core Processor
Cheng, Xuntao
He, Bingsheng
Lu, Mian
Lau, Chiew Tong
Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
description Emerging many-core processors feature very high memory bandwidth and computational power. For example, Intel Xeon Phi many-core processors of the Knights Corner (KNC) and Knights Landing (KNL) architectures embrace 60 to 64 x86-based CPU cores with 512-bit SIMD capabilities and high-bandwidth memories like the GDDR5 on KNC and on-package DRAMs on KNL. In this paper, we study the performance main-memory database operators and online analytical processing (OLAP) on such many-core architectures. We find that even the state-of-the-art database operators suffer severely from memory stalls and resource underutilization on those many-core processors. We argue that a software approach decomposing a coarse-grained operator into fine-grained phases and executing two independent phases with complementary resource requirements concurrently can address this problem. This approach allows more fine-grained control of resource utilization. Our experiments demonstrate significant performance gain and high resource utilization achieved by our proposed approaches on both KNC and KNL.
author2 Interdisciplinary Graduate School (IGS)
author_facet Interdisciplinary Graduate School (IGS)
Cheng, Xuntao
He, Bingsheng
Lu, Mian
Lau, Chiew Tong
format Article
author Cheng, Xuntao
He, Bingsheng
Lu, Mian
Lau, Chiew Tong
author_sort Cheng, Xuntao
title Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
title_short Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
title_full Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
title_fullStr Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
title_full_unstemmed Many-core needs fine-grained scheduling : a case study of query processing on Intel Xeon Phi processors
title_sort many-core needs fine-grained scheduling : a case study of query processing on intel xeon phi processors
publishDate 2020
url https://hdl.handle.net/10356/141470
_version_ 1681056803060711424