AMNES: Accelerating the Computation of Data Correlation Using FPGAs

Author:

Chiosa Monica1,Preußer Thomas B.2,Blott Michaela2,Alonso Gustavo1

Affiliation:

1. Systems Group, Dept. of Computer Science, ETH Zurich

2. AMD Research and Advanced Development

Abstract

A widely used approach to characterize input data in both databases and ML is computing the correlation between attributes. The operation is supported by all major database engines and ML platforms. However, it is an expensive operation as the number of attributes involved grows. To address the issue, in this paper we introduce AMNES, a stream analytics system offloading the correlation operator into an FPGA-based network interface card. AMNES processes data at network line rate and the design can be used in combination with smart storage or SmartNICs to implement near data or in-network data processing. AMNES design goes beyond matrix multiplication and offers a customized solution for correlation computation bypassing the CPU. Our experiments show that AMNES can sustain streams arriving at 100 Gbps over an RDMA network, while requiring only ten milliseconds to compute the correlation coefficients among 64 streams, an order of magnitude better than competing CPU or GPU designs.

Publisher

Association for Computing Machinery (ACM)

Subject

General Earth and Planetary Sciences,Water Science and Technology,Geography, Planning and Development

Reference85 articles.

1. Statology 2021. Correlation Does Not Imply Causation: 5 Real-World Examples. Statology. https://www.statology.org/correlation-does-not-imply-causation-examples/ Statology 2021. Correlation Does Not Imply Causation: 5 Real-World Examples. Statology. https://www.statology.org/correlation-does-not-imply-causation-examples/

2. uBFT: Microsecond-Scale BFT using Disaggregated Memory

3. An ultra-high-speed FPGA based digital correlation processor

4. AMD-Xilinx. [n.d.]. Floating-Point Operator v7. 1 Product Guide (PG060) . Retrieved May 17, 2023 from https://docs.xilinx.com/v/u/en-US/pg060-floating-point AMD-Xilinx. [n.d.]. Floating-Point Operator v7.1 Product Guide (PG060). Retrieved May 17, 2023 from https://docs.xilinx.com/v/u/en-US/pg060-floating-point

Cited by 1 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3