Dataset Information

Evaluating architecture impact on system energy efficiency.

ABSTRACT: As the energy consumption has been surging in an unsustainable way, it is important to understand the impact of existing architecture designs from energy efficiency perspective, which is especially valuable for High Performance Computing (HPC) and datacenter environment hosting tens of thousands of servers. One obstacle hindering the advance of comprehensive evaluation on energy efficiency is the deficient power measuring approach. Most of the energy study relies on either external power meters or power models, both of these two methods contain intrinsic drawbacks in their practical adoption and measuring accuracy. Fortunately, the advent of Intel Running Average Power Limit (RAPL) interfaces has promoted the power measurement ability into next level, with higher accuracy and finer time resolution. Therefore, we argue it is the exact time to conduct an in-depth evaluation of the existing architecture designs to understand their impact on system energy efficiency. In this paper, we leverage representative benchmark suites including serial and parallel workloads from diverse domains to evaluate the architecture features such as Non Uniform Memory Access (NUMA), Simultaneous Multithreading (SMT) and Turbo Boost. The energy is tracked at subcomponent level such as Central Processing Unit (CPU) cores, uncore components and Dynamic Random-Access Memory (DRAM) through exploiting the power measurement ability exposed by RAPL. The experiments reveal non-intuitive results: 1) the mismatch between local compute and remote memory node caused by NUMA effect not only generates dramatic power and energy surge but also deteriorates the energy efficiency significantly; 2) for multithreaded application such as the Princeton Application Repository for Shared-Memory Computers (PARSEC), most of the workloads benefit a notable increase of energy efficiency using SMT, with more than 40% decline in average power consumption; 3) Turbo Boost is effective to accelerate the workload execution and further preserve the energy, however it may not be applicable on system with tight power budget.

SUBMITTER: Yu S

PROVIDER: S-EPMC5697812 | biostudies-literature | 2017

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Evaluating architecture impact on system energy efficiency.

Yu Shijie S Yang Hailong H Wang Rui R Luan Zhongzhi Z Qian Depei D

PloS one 20171121 11

As the energy consumption has been surging in an unsustainable way, it is important to understand the impact of existing architecture designs from energy efficiency perspective, which is especially valuable for High Performance Computing (HPC) and datacenter environment hosting tens of thousands of servers. One obstacle hindering the advance of comprehensive evaluation on energy efficiency is the deficient power measuring approach. Most of the energy study relies on either external power meters ...[more]

PMID: 29161317

Similar Datasets

Project description:Goal-directed therapy guidelines for pediatric septic shock resuscitation recommend fluid delivery at speeds in excess of that possible through use of regular fluid infusion pumps. In our experience, syringes are commonly used by health care providers (HCPs) to achieve rapid fluid resuscitation in a pediatric fluid resuscitation scenario. At present, it is unclear which syringe size health care providers should use when performing fluid resuscitation to achieve maximal fluid resuscitation efficiency. The objective of this study was therefore to determine if an optimal syringe size exists for conducting manual pediatric fluid resuscitation.This 48-participant parallel group randomized controlled trial included 4 study arms (10, 20, 30, 60 mL syringe size groups). Eligible participants were HCPs from McMaster Children's Hospital, Hamilton, Canada blinded to the purpose of the trial. Consenting participants were randomized using a third party technique. Following a standardization procedure, participants administered 900 mL (60 mL/kg) of isotonic saline to a simulated 15 kg child using prefilled provided syringes of the allocated size in rapid sequence. Primary outcome was total time to administer the 900 mL and this data was collected through video review by two blinded outcome assessors. Sample size was predetermined based upon a primary outcome analysis using one-way ANOVA.12 participants were randomized to each group (n=48) and all completed trial protocol to analysis. Analysis was conducted according to intention to treat principles. A significant difference in fluid resuscitation time (in seconds) was found between syringe size group means: 10 mL, 563s [95% CI 521; 606]; 20 mL, 506s [95% CI 64; 548]; 30 mL, 454s [95% CI 412; 596]; 60 mL, 455s [95% CI 413; 497] (p<0.001).The syringe size used when performing manual pediatric fluid resuscitation has a significant impact on fluid resuscitation speed, in a setting where fluid filled syringes are continuously available. Greatest efficiency was achieved with 30 or 60 mL syringes.ClinicalTrials.gov, NCT01494116.

Dataset Information

Evaluating architecture impact on system energy efficiency.

Publications

Evaluating architecture impact on system energy efficiency.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure