A PCIe DMA engine to support the virtualization of 40 Gbps FPGA-accelerated network appliances

Abstract

Network Function Virtualization (NFV) allows creating specialized network appliances out of general-purpose computing equipment (servers, storage, and switches). In this paper we present a PCIe DMA engine that allows boosting the performance of virtual network appliances by using FPGA accelerators. Two key technologies are demonstrated, SR-IOV and PCI Passthrough. Using these two technologies, a single FPGA board can accelerate several virtual software appliances. The final goal is, in an NFV scenario, to substitute conventional Ethernet NICs by networking FPGA boards (such as NetFPGA SUME). The advantage of this approach is that FPGAs can very efficiently implement many networking tasks, thus boosting the performance of virtual networking appliances. The SR-IOV capable PCIe DMA engine presented in this work, as well as its associated driver, are key elements in achieving this goal of using FPGA networking boards instead of conventional NICs. Both DMA engine and driver are open source, and target the Xilinx 7-Series and UltraScale PCIe Gen3 endpoint. The design has been tested on a NetFPGA SUME board, offering transfer rates reaching 50 Gb/s for bulk transmissions. By taking advantage of SR-IOV and PCI Passthrough technologies, our DMA engine provides transfers rate well above 40 Gb/s for data transmissions from the FPGA to a virtual machine. We have also identified the bottlenecks in the use of virtualized FPGA accelerators caused by reductions in the maximum read request size and maximum payload PCIe parameters. Finally, the DMA engine presented in this paper is a very compact design, using just 2% of a Xilinx Virtex-7 XC7VX690T device.

Veritabanı Uygulamalarında FPGA Tabanlı Hızlandırıcı Kullanımı

“…Bunun verimli çalışabilmesi için, DMA (Direct Memory Access) denilen yapının kullanılması gerekmektedir. Bu sayede, veri DRAM tabanlı hafıza seviyesinden direk olarak FPGA (Field Programmable Gate Array) yapısına aktarılaracak, hafıza hiyerarşisinde herhangi bir etkisi olmayacaktır (Zazo, Lopez-Buedo, Audzevich, & Moore, 2015).…”

Performance assessment of 40 Gbit/s off-the-shelf network cards for virtual network probes in 5G networks

“…A similar performance limitation in network virtualization is discussed in [30], where DPDK is used to increase the throughput of Open vSwitch. Specialized hardware, such as FPGAs [31], could also be used to increase the performance of virtual networks, although the issues regarding cost and scalability that discussed previously would remain in this scenario.…”

Temporal Accelerators: Unleashing the Potential of Embedded FPGAs

“…FPGAs can be used to execute tailormade accelerators to efficiently perform complex calculations "in hardware", while having the capability to change the circuit when required. FPGA-based accelerators have been shown to increase the overall performance and efficiency significantly of applications in fields such as artificial intelligence (AI) and deep learning [Biswas et al 2018, Han et al 2015, Roth et al 2020, networking [Varga et al 2015, Zazo et al 2015, Ruiz et al 2019, cryptography [Chelton et al 2008, Aysu et al 2013, Koziel et al 2016, Babaei et al 2019, database systems [Ziener et al 2016] or control systems [Monmasson et al 2007, Montealegre et al 2015.…”