Abstract-High Performance Computing (HPC) data centers are becoming increasingly dense; the associated power-density and energy consumption of their operation is increasing. Up to half of the total energy is attributed to cooling the data center; greening the data center operations to reduce both computing and cooling energy is imperative. To this effect: i) the Energy Inefficiency Ratio of SPatial job scheduling (a.k.a. job placement) algorithms, also referred as SP-EIR, is analyzed by comparing the total (computing + cooling) energy consumption incurred by the algorithms with the minimum possible energy consumption, while assuming that the job start times are already decided to meet the Service Level Agreements (SLAs); and ii) a coordinated coolingaware job placement and cooling management algorithm, Highest Thermostat Setting (HTS), is developed. HTS is aware of dynamic behavior of the Computer Room Air Conditioner (CRAC) units and places the jobs in a way to reduce the cooling demands from the CRACs. Dynamic updates of the CRAC thermostat settings based on the cooling demands can enable a reduction in energy consumption. Simulation results based on power measurements and job traces from the ASU HPC data center show that: i) HTS reduces the SP-EIR by 15% compared to LRH, a thermal-aware spatial scheduling algorithm; and ii) in conjunction with FCFSBackfill, HTS increases the throughput per unit energy by 6.89% and 5.56%, respectively, over LRH and MTDP (an energy-efficient spatial scheduling algorithm with server consolidation).
I. IHigh Performance Computing (HPC) applications require high computation capabilities, often in the range of teraflops. A major issue in contemporary data centers, hosting such high computation facilities, is the high energy consumption in their operations. Indeed, the data centers' energy consumption amounted to nearly 2% of the total energy budget of the US in 2007 and is expected to reach 4% in 2011 [1]; as such, greening the data center operations has been of utmost interest over the years [2]-[8]. Up to half of this energy can be attributed to cooling the data centers (i.e. cooling energy) to keep the operating temperatures within manufacturer specified redline temperatures. This paper focuses on a cyber-physical oriented coordinated job and cooling management in HPC data centers to reduce the total (i.e. computing and cooling) energy consumption of the data centers.The cooling energy depends on two factors: i) the cooling demand, which is driven by the power distribution and the redline temperature; and ii) the cooling behavior, i.e. the behavior of the Computer Room Air Conditioner (CRAC) unit (controlled by varying the thermostat setting), to meet the demand. A major concern in this regard is the possible recirculation and intermixing of hot air generated by running the jobs with the cold air supplied from the CRAC [7]. Recirculation of hot air depends on the data center layout and can cause hot-spots; thus potentially increasing the cooling demand.Techniques to r...
Galaxy clusters have the potential to accelerate cosmic rays (CRs) to ultrahigh energies via accretion shocks or embedded CR acceleration sites. The CRs with energies below the Hillas condition will be confined within the cluster and eventually interact with the intracluster medium gas to produce secondary neutrinos and gamma rays. Using 9.5 yr of muon neutrino track events from the IceCube Neutrino Observatory, we report the results of a stacking analysis of 1094 galaxy clusters with masses ≳1014
M
⊙ and redshifts between 0.01 and ∼1 detected by the Planck mission via the Sunyaev–Zel’dovich effect. We find no evidence for significant neutrino emission and report upper limits on the cumulative unresolved neutrino flux from massive galaxy clusters after accounting for the completeness of the catalog up to a redshift of 2, assuming three different weighting scenarios for the stacking and three different power-law spectra. Weighting the sources according to mass and distance, we set upper limits at a 90% confidence level that constrain the flux of neutrinos from massive galaxy clusters (≳1014
M
⊙) to be no more than 4.6% of the diffuse IceCube observations at 100 TeV, assuming an unbroken E
−2.5 power-law spectrum.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.