Thermal Performance of a Passive Immersion-Cooling Multichip Module

1995 ◽  
Vol 2 (1-2) ◽  
pp. 95-103 ◽  
Author(s):  
David Kitching ◽  
Tetsuo Ogata ◽  
Avram Bar-Cohen
Author(s):  
Shuai Shao ◽  
Tianyi Gao ◽  
Huawei Yang ◽  
Jie Zhao ◽  
Jiajun Zhang

Abstract Along with advancements in microelectronics packaging, the power density of processor units has steadily increased over time. Data center servers equipped for high performance computing (HPC) often use multiple central processing units (CPUs) and graphical processing units (GPUs), thereby resulting in an increased power density, exceeding 1 kW per U. Many data center organizations are evaluating single phase immersion technology as a potential energy and resource saving cooling option. In this work immersion cooling was studied at a power level of 2.7kW/U with a 5U-height immersion cooling tank. Heat generated by a simulated GPU server was transferred to the secondary loop coolant, and then exchanged with the primary loop facility coolant through the heat exchanger. The chiller supply and return temperature and flow rate was controlled for the primary loop. The simulated GPU server chassis was designed to provide thermal power equivalent to a high power density server. Eight simulated power heaters, of which each unit was the size of a GPU chipset, was assembled in the comparable location to a real IT equipment on a 4U server chassis. Power for the GPU simulated chassis was able to support up to 2700 W maximum. Three investigations for this immersion cooling system evaluation were performed through comprehensive testing. The first is to identify the key decision making factor(s) for evaluating the thermal performance of 4 hydrocarbon-based dielectric coolants, including power parametric analysis, transient analysis, power cycling test, and fluid temperature profiling. The second is to develop an optimization strategy for the immersion system thermal performance. The third is to verify the capability of an 1U heat sink to support high density processor units over 300 W per GPU in an immersion cooling solution.


2011 ◽  
Vol 2011 (1) ◽  
pp. 000683-000689
Author(s):  
Ranjith John ◽  
Vladimir Dotsenko ◽  
Deepnarayan Gupta ◽  
Ajay Malshe

We report the experimental study of the thermal resistance of a flip chip bonded superconducting multichip module (MCM) in a liquid cryogen free environment. A 5×5 mm2 indium-tin bumped superconducting chip was flip chip bonded on a 1×1 cm2 superconducting carrier chip. A non-conductive adhesive was used as an underfill to enhance the robustness of the package. We designed a test bed where the LSCE module was mounted onto the cold head of a Gifford McMahon (GM) cryocooler. The module was conductively cooled down to 4 K and the thermal resistance between the chip and the carrier chip was analyzed. The experimental results showed that for the power dissipation (2 – 5 mW), which is typical for low temperature superconducting electronic LSCE devices, the thermal resistance was 20.1 +/− 1.9 K/W. Thermal model of the current LSCE package was investigated using COMSOL multi-physics. Theoretical estimates showed that for the current package setup the expected thermal resistance of the bump path to be 6.2 K/W. The discrepancy between the model and experimental analysis has been explained due to the presence of voids and inadequate bump contact area. To our knowledge, this is the first such experimental investigation of the thermal performance of adhesive bonded LSCE package on a cryocooler. This experimental analysis is of paramount importance for future trends in single chip and multichip module packaging of LSCE devices.


Author(s):  
Pravin A. Shinde ◽  
Pratik V. Bansode ◽  
Satyam Saini ◽  
Rajesh Kasukurthy ◽  
Tushar Chauhan ◽  
...  

Abstract Liquid immersion cooling of servers in synthetic dielectric fluids is an emerging technology which offers significant cooling energy savings and increased power densities for data centers. A noteworthy advantage of using immersion cooling is high heat dissipation capacity which is roughly 1200 times greater than air. Other advantages of dielectric fluid immersion cooling include high rack density, better server performance, even temperature profile, reduction in noise etc. The enhanced thermal properties of oil lead to the considerable savings in both upfront and operating cost over traditional methods. In this study, a server is completely submerged in a synthetic dielectric fluid. Experiments are conducted to observe the effects of varying the volumetric flow rate and oil inlet temperature on thermal performance and power consumption of the server. Various parameters like total server power consumption, the temperature of all heat generating components like Central Processing Unit (CPU), Dual in Line Memory Module (DIMM), input/output hub (IOH) chip, Platform Controller Hub (PCH), Network Interface Controller (NIC) are measured at steady state. Since this is an air-cooled server, the results obtained from the experiments will help in proposing better heat removal strategies like heat sink optimization, better ducting and server architecture. Assessment has been made on the effect of thermal shadowing caused by the two CPUs on the nearby components like DIMMs and PCH.


Author(s):  
Dhruvkumar Gandhi ◽  
Uschas Chowdhury ◽  
Tushar Chauhan ◽  
Pratik Bansode ◽  
Satyam Saini ◽  
...  

Abstract Complete immersion of servers in synthetic dielectric fluids is rapidly becoming a popular technique to minimize the energy consumed by data centers for cooling purposes. In general, immersion cooling offers noteworthy advantages over conventional air-cooling methods as synthetic dielectric fluids have high heat dissipation capacities which are roughly about 1200 times greater than air. Other advantages of dielectric fluid immersion cooling include even thermal profile on chips, reduction in noise and addressing reliability and operational enhancements like whisker formation and electrochemical migration. Nevertheless, lack of data published and availability of long-term reliability data on immersion cooling is insufficient which makes most of data centers operators reluctant to implement this technique. The first part of this paper will compare thermal performance of single-phase oil immersion cooled HP ProLiant DL160 G6 server against air cooled server using computational fluid dynamics on 6SigmaET®. Focus of the study are major components of the server like Central Processing Unit (CPU), Dual in Line Memory Module (DIMM), Input/output Hub (IOH) chip and Input/output controller Hub (ICH). The second part of this paper focuses on thermal performance optimization of oil immersion cooled servers by varying inlet oil temperature, flow rate and using different fluid.


Sign in / Sign up

Export Citation Format

Share Document