How beneficial is peer-to-peer DMA?

How beneficial is peer-to-peer DMA?
复制标题

点对点 DMA 有多大好处?

DOI:
10.1145/3409963.3410491
复制
发表时间:
2020
期刊:
APSys '20: Proceedings of the 11th ACM SIGOPS Asia-Pacific Workshop on Systems
影响因子:
--
通讯作者:
Akashi Kunio
Akashi Kunio
中科院分区:
--
文献类型:
--
作者:
Nakamura Ryo;Kuga Yohei;Akashi Kunio

文献摘要

参考文献

被引文献

相似文献

本文重新审视对等DMA(P2P DMA),并研究其在以太网SSD和NVMe SSD上的开发潜力。CPU性能提升的放缓导致了外围加速器的出现,如Smart Pentium和TPU。P2P DMA通过避免数据在主存储器上反弹,为多个外围设备的有效集成提供了潜力。然而,P2P DMA主要是围绕GPU进行研究的,并且其改进已经针对特定应用进行了测量。在本文中,我们进行了实验,以澄清使用P2P DMA对个别设备的好处,即,以太网NIC和NVMe SSD,从I/O吞吐量的角度来看。我们开发了一个名为Libpop的库,用于操纵设备上的内存以调用P2P DMA。此外,我们还将Libpop集成到pcie-bench中,pcie-bench是一款基于FPGA的基准设备,netmap用于以太网以太网,UNVMe用于NVMe SSD。这些实现的实验表明:(1)内存写入会使DMA写入的吞吐量降低70%,(2)这种降低会影响设备上的I/O吞吐量,(3)P2P DMA可以避免降低,但设备队列会影响以太网NIC上的吞吐量。
This paper revisits Peer-to-Peer DMA (P2P DMA) and investigates its potential for exploitation on Ethernet NICs and NVMe SSDs. The slowing performance improvement of CPUs has led to emergence of peripheral accelerators such as Smart NICs and TPU. P2P DMA presents potential for the efficient integration of multiple peripherals by avoiding data bouncing on main memory. However, P2P DMA has been studied mainly around GPUs, and its improvement has been measured for specific applications. In this paper, we perform experiments to clarify the benefits of using P2P DMA on individual devices, i.e., an Ethernet NIC and an NVMe SSD, from an I/O throughput perspective. We developed a library, called Libpop, for manipulating memory on devices for invoking P2P DMA. Additionally, we integrated Libpop into pcie-bench, which is an FPGA-based benchmark device, netmap for Ethernet NICs, and UNVMe for NVMe SSDs. Experiments with these implementations show that (1) memory writes degrade the throughput of DMA write by 70%, (2) the degradation affects I/O throughput on the devices, and (3) P2P DMA can avoid degradation, but device queues affect throughput on the Ethernet NIC.
DOI: 10.1109/cloudcom.2017.14
发表时间: 2017-12
期刊: 2017 IEEE International Conference on Cloud Computing Technology and Science (CloudCom)
影响因子: --
作者:
Ziye Yang;James R. Harris;Benjamin Walker;Daniel Verkamp;Changpeng Liu;Cunyin Chang;Gang Cao;J. Stern;Vishal Verma;Luse E. Paul
通讯作者: Ziye Yang;James R. Harris;Benjamin Walker;Daniel Verkamp;Changpeng Liu;Cunyin Chang;Gang Cao;J. Stern;Vishal Verma;Luse E. Paul