0
Guest Editorial

Introduction for Journal of Electronic Packaging Special Section on Data Centers PUBLIC ACCESS

J. Electron. Packag 135(3), 030301 (Jul 24, 2013) (1 page) Paper No: EP-13-1067; doi: 10.1115/1.4024993 History: Received July 11, 2013; Revised July 11, 2013
FIGURES IN THIS ARTICLE

As the digital factories of the information technology (IT) age, data centers are showing an unprecedented growth. From 2012 to 2016 alone, the global data center traffic is expected to increase four fold [1,2-1,2], requiring new facilities to be built at the compound annual growth rate of 15.78% during the same period, with the associated positive economic impact. These beneficial outcomes are tempered by concerns about sustainability of such explosive growth, in particular, the energy and water usage burden of the new facilities. Larger facilities can consume tens of megawatts of electric power, with 20–50% of it used for cooling. As such, energy efficient thermal management of current and future facilities has emerged as a significant research challenge and is the focus of this Special Section.

Abdelmaksoud, Dang, Khalifa, and Schmidt (Improved computational fluid dynamics (CFD) Model for Open-Aisle Air-Cooled Data Center Simulations) present CFD simulations for data center air flows using modifications to the floor tile boundary conditions, which result in improved agreement with experimental results. It is shown that fully open tile boundary condition underspecifies the initial jet momentum, resulting in significant error in flow predictions. Air flow modeling for a single rack in a data center is explored by Arghode, Kumar, Weiss, Meyer, and Joshi (Rack Level Modeling of Air Flow through Perforated Tile in a Data Center), with a focus on the effect of the modeling of perforated floor tiles. It is found that agreement with particle image velocimetry measurements improves, as geometric details, such as pore locations and shapes are included. A modification of the body force method to account for the momentum increase across the pores is also presented.

CFD simulations of convective transport in data centers are computationally intensive, and are as such often unsuitable for parametric design optimization, and real time control of cooling equipment. Lettieri, Toulouse, Shah, Bash, and Carey (Computational and Experimental Validation of a Vortex-Superposition-Based Buoyancy Approximation for the Compact Model of Potential Flow and Convective Transport (COMPACT) Code in Data Centers) present a modification to their rapid modeling COMPACT code to include the effects of buoyancy, using a vortex superposition method. They also compare the predictions with temperature measurements, showing an improvement in the predictive capability of COMPACT. Cruz and Joshi (Inviscid and Viscous Numerical Models Compared to Experimental Data in a Small Data Center Test Cell) present a comparison of laminar, and several Reynolds averaged Navier Stokes based turbulent modeling approaches, with inviscid models of a data center cell. The inviscid model ran over 30% faster than the fastest viscous model, and also provided the best agreement with experimental measurements, a finding which the authors suggest may be dependent on the particular data center layout evaluated.

Thermal modeling of data centers is useful for both design of data centers, as well as for co-optimization of energy usage for IT and thermal management. Two optimization approaches are explored by Behnia, Fakhim, Armfield, and Nagarathinam (A Comparison of Parametric and Multi-Variable Optimization Techniques in a Raised-Floor Data Centre) to optimize the layout of a data center. A parametric optimization approach is used to optimize the maximum air temperature within the room. A second multivariable approach, which minimizes a composite cost function is found to be superior in terms of computational resource use and thermal performance. A proper orthogonal decomposition (POD) based reduced order modeling approach is used by Demetriou and Khalifa (Thermally Aware, Energy-Based Load Placement in Open-Aisle, Air-Cooled Data Centers) to predict rack inlet air temperatures. The POD model requires a set of CFD solutions, and can generate subsequent predictions rapidly, guiding IT workload placement.

As data centers continue to house hardware with increasing functionality and performance, the rack level heat removal requirements are rapidly increasing, prompting the need to look at liquid cooling approaches. Lamaison, Marcinichen, and Thome (Two-Phase Flow Control of Electronics Cooling with Pseudo-CPUs in Parallel Flow Circuits: Transient Modeling and Experimental Evaluation) investigate two-phase cooling of multiple chips. They focus on managing spatially nonuniform and time varying power dissipations, while avoiding dry out. A transient modeling and control framework was developed and validated with experiments conducted with a dual evaporator test-bed. The predictive framework was also utilized to simulate the two-phase cooling of four microprocessors in parallel.

References

Cisco Systems Inc., 2012, “Cisco Global Cloud Index Forecasts Cloud Traffic to Grow Sixfold by 2016,” http://newsroom.cisco.com/release/1091855
Infiniti Research Ltd., 2013, “Global Data Center Construction Market 2012-2016,” http://www.researchandmarkets.com/research/tkz5pc/global_data
Copyright © 2013 by ASME
This article is only available in the PDF format.

References

Cisco Systems Inc., 2012, “Cisco Global Cloud Index Forecasts Cloud Traffic to Grow Sixfold by 2016,” http://newsroom.cisco.com/release/1091855
Infiniti Research Ltd., 2013, “Global Data Center Construction Market 2012-2016,” http://www.researchandmarkets.com/research/tkz5pc/global_data

Figures

Tables

Errata

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In