Power shifting in Thrifty Interconnection Network

Li, Jian; Huang, Wei; Lefurgy, Charles; Zhang, Lixin; Denzel, Wolfgang E.; Treumann, Richard; Wang, Kun

doi:10.1109/hpca.2011.5749725

Cited by 23 publications

(19 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Jian Li et al, [17] proposed a similar on/off based interconnect, which makes use of system events to overlap link on transaction delay. Their approach requires a control network which sends messages that switch on the links prior to sending the message and thereby overlaps the link on transaction delay with the pre-processing of messages.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Power/performance evaluation of energy efficient Ethernet (EEE) for High Performance Computing

Saravanan

Carpenter

Ramirez

2013

2013 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)

View full text Add to dashboard Cite

As of June 2012, 41 % of all systems in the TOPSOO use Gigabit Ethernet. Ethernet has been a strong contender in the HPC interconnect market for its competitive performance and low cost. However, until recently, little emphasis has been thrown upon bringing about energy efficient HPC interconnects.To illustrate, in a majority if not all Ethernet based systems, the transmitter and receiver operate at full power regardless of any data transmission between them, leading to power inefficiency.The recent standard IEEE 802.3az, Energy Efficient Ethernet (EEE), approved in 2010, solves the above conundrum by intro ducing "Low-Power-Idle", dynamically turning off unused links to save interconnect power.In this paper, we present the first analysis of Energy Efficient Ethernet in the domain of HPC, examining its potential for power savings. Unlike previous proposals, we present a detailed analysis of the impact of additional latency overhead introduced by EEE, using multiple simulated systems running actual HPC application traces. We propose the use of "Power-Down Threshold", as a possible add-on to EEE to mitigate its on/off transition overhead.We find that EEE brings about link power savings of about 70% by switching off links, but at the cost of performance, leading to increased power consumption of the overall system by lS% (average). In contrast, using our proposed "Power-DownThreshold", we demonstrate reduced on/off transition overhead, from 2S% to 2%, translating to overall system power savings of about 7.S%. Furthermore, in this work we point out relevant design decisions for future vendors intending to deploy EEE solutions for their HPC systems.

show abstract

Section: Related Workmentioning

confidence: 99%

“…To model power consumption, we use data available in the industry and relevant literature [7,8,11,14,17]. We model the power consumption of links to consume 10% of the full link power in idle mode based on the IEEE 802.3az EEE [1] draft.…”

Section: For Our Analysis Low Power Idle Mode Of Energy Efficientmentioning

confidence: 99%

Power/performance evaluation of energy efficient Ethernet (EEE) for High Performance Computing

Saravanan

Carpenter

Ramirez

2013

2013 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)

View full text Add to dashboard Cite

show abstract

“…Jian et al [24] work is focused on non-prediction powersaving techniques. Links are powered up just before they are needed, by relaying on hints from the built-in system events or from macros in MPI source code.…”

Section: Related Workmentioning

confidence: 99%

Software-Managed Power Reduction in Infiniband Links

Dickov

Pericàs

Carpenter

et al. 2014

2014 43rd International Conference on Parallel Processing

View full text Add to dashboard Cite

Abstract-Interconnection network represents the backbone of large-scale supercomputers. While the computation has become more energy proportional, the interconnection network has started to become one of the important consumers in the system. Still, interconnection network has received little attention in terms of efficient power consumption. Several hardware-based schemes have been proposed that function either by turning off the links or lowering the voltage/frequency in links when idle or with low usage. Although effective in certain cases, these schemes share the common drawback that they do not have enough global information about the application to manage network power efficiently. This paper, instead, proposes a software prediction support for shutting down inactive lanes of an Infiniband (IB) link.Using n-gram extraction technique as a core of the prediction algorithm we are able to predict with high accuracy when links will be unused. The prediction algorithm detects the repeating patterns of MPI calls during the program execution, on which basis the prediction of the following MPI calls can be realized with relatively high precision. This allow us to know on time when links will be active/idle and invoke the explicit turn-off/on lanes calls. Our power saving mechanism is adopted to be executed within MPI allowing a large number of existing MPI programs to be run without any source code modification. Results show an average power savings in IB switches up to around 33% over the representative HPC applications. On the other hand, HPC applications average suffer only up to 1% increase in execution time.

show abstract

“…Among hardware based approaches, power management of interconnection networks using on/off links has been studied [3,18,19]. On/off links, which refers to shutting down communication links that are not being used, has shown to be a useful method to save power.…”

Section: A Related Workmentioning

confidence: 99%

Toward Runtime Power Management of Exascale Networks by on/off Control of Links

Totoni

Jain

Kalé

2013

2013 IEEE International Symposium on Parallel &Amp; Distributed Processing, Workshops and PHD Forum

View full text Add to dashboard Cite

Abstract-Higher radix networks, such as high-dimensional tori and multi-level directly connected networks, are being used for supercomputers as they become larger but need lower diameter. These networks have more resources (e.g. links) in order to provide good performance for a range of applications. We observe that a sizeable fraction of the links in the interconnect are never used or underutilized during execution of common parallel applications. Thus, in order to save power, we propose addition of hardware support for on/off control of links in software and their management using adaptive runtime systems.We study the effectiveness of our approach using real applications (NAMD, MILC), and application benchmarks (NAS Parallel Benchmarks, Jacobi). They are simulated on representative networks such as 6-D Torus and IBM PERCS (similar to Dragonfly). For common applications, our approach can save up to 16% of total machine's power and energy, without any performance penalty.

show abstract

Power shifting in Thrifty Interconnection Network

Cited by 23 publications

References 22 publications

Power/performance evaluation of energy efficient Ethernet (EEE) for High Performance Computing

Power/performance evaluation of energy efficient Ethernet (EEE) for High Performance Computing

Software-Managed Power Reduction in Infiniband Links

Toward Runtime Power Management of Exascale Networks by on/off Control of Links

Contact Info

Product

Resources

About