Power-efficient networking for balanced system designs

Byrne, John; Chang, Jichuan; Lim, Kevin; Ramírez, Laura; Ranganathan, Parthasarathy

doi:10.1145/2039252.2039255

Cited by 10 publications

(4 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Comparing with these interconnection specifications, PCIe has advantageous characteristics such as higher protocol efficiency, low-power consumption per port, and low unit price [1], which makes developers endeavor to use PCIe as system interconnects in small-scale clusters [2]- [3].…”

Section: Using Pci Express For System Interconnectsmentioning

confidence: 99%

Implementation of system interconnection devices using PCI express

Ren

Kim

2015

2015 IEEE International Conference on Consumer Electronics (ICCE)

View full text Add to dashboard Cite

PCI Express is considered as a good candidate for system interconnects due to its fast speed, low power consumption, and good protocol efficiency. In this paper we propose a system interconnection device based on PCI Express for usage in small-scale system interconnect and describe its design, implementation, and early evaluation. The proposed device provides multiple 20 Gbps connections and current testing results show that the throughput reaches maximally 6.797 Gbps without DMA.

show abstract

Section: Using Pci Express For System Interconnectsmentioning

confidence: 99%

Implementation of system interconnection devices using PCI express

Ren

Kim

2015

2015 IEEE International Conference on Consumer Electronics (ICCE)

View full text Add to dashboard Cite

show abstract

“…Unfortunately, due to lack of MRA switch in the market, many proprietary attempts have been made to achieve multi-root sharing using SRIOV compliant I/O devices with customized hardware boxes [4,7,10,28,44] or relying on NTB [32,40,41] as an alternative. Besides, PCIe could be an ideal power-efficient interconnect for intra-rack communications [5,20,26], by mapping parts of the client host's memory to a global memory address space [23,24]. As a point of comparison, Marlin maps the hosts' entire memory address space to avoid extra data copying and guarantees security by leveraging existing hardware components.…”

Section: Related Workmentioning

confidence: 99%

Marlin

Lee

Chiueh

2014

Proceedings of the Tenth ACM/IEEE Symposium on Architectures for Networking and Communications Systems

View full text Add to dashboard Cite

Disaggregation of hardware resources that are traditionally embedded within individual servers into separate resource pools is an emerging architectural trend in hyperscale data center design, as exemplified by Facebook's disaggregated rack architecture. This paper presents the design, implementation and evaluation of a PCIebased rack area network system called Marlin, which is designed to support the communications and resource sharing needs of disaggregated racks. By virtue of being based on PCIe, Marlin presents a memory-based addressing model for both I/O device sharing among multiple hosts and inter-host communications. That is, when a node communicates with other nodes or accesses resources in the same rack, it uses memory read and write operations. In the area of inter-node communications, Marlin offers hardware-based remote direct memory access (HRDMA) as a first-class communications primitive between servers within a rack. In addition, Marlin supports socket-based communications for legacy network applications and cross-machine zero memory copying for applications designed specifically to take full advantage of memory-based communications. Empirical measurements on a fully operational Marlin prototype based on 4-lane Gen3 PCIe technology show that the one-way kernel-to-kernel latency is 8.5µsec and the end-to-end sustainable TCP throughput is 19.6 Gbps.

show abstract

“…With proper software, it is even possible for a cluster node to leverage multiple NICs simultaneously. Third, the PCIe network connecting the cluster nodes and the I/O devices they share could substantially reduce the power consumption of intra-cluster communications because the PCIe transceiver is meant for short-distance transmission and thus is much simpler and less power-hungry than 10GE or Infiniband [7]. Coupling I/O device sharing with I/O virtualization enables multiple VMs on multiple servers to interact directly with PCIe devices shared among the servers.…”

Section: Introductionmentioning

confidence: 99%

Secure I/O device sharing among virtual machines on multiple hosts

Lee

Chiueh

2013

Proceedings of the 40th Annual International Symposium on Computer Architecture

View full text Add to dashboard Cite

Virtualization allows flexible mappings between physical resources and virtual entities, and improves allocation efficiency and agility. Unfortunately, most existing virtualization technologies are limited to resources in a single host. This paper presents the design, implementation and evaluation of a multi-host I/O device virtualization system called Ladon, which enables I/O devices to be shared among virtual machines running on multiple hosts in a secure and efficient way. Specifically, Ladon uses a PCIe network to connect multiple servers with PCIe devices and allows VMs running on these servers to directly interact with these PCIe devices without interfering with one another. Through an evaluation of a fully operational Ladon prototype, we show that there is no throughput and latency penalty of the multi-host I/O virtualization enabled by Ladon compared to those of the existing single-host I/O virtualization technology.

show abstract

Power-efficient networking for balanced system designs

Cited by 10 publications

References 12 publications

Implementation of system interconnection devices using PCI express

Implementation of system interconnection devices using PCI express

Marlin

Secure I/O device sharing among virtual machines on multiple hosts

Contact Info

Product

Resources

About