Enabling Thermal Management of High-Powered Server Processors Using Passive Thermosiphon Heat Sink

Author(s):  
Devdatta P. Kulkarni ◽  
Priyanka Tunuguntla ◽  
Guixiang Tan ◽  
Casey Carte

Abstract In recent years, rapid growth is seen in computer and server processors in terms of thermal design power (TDP) envelope. This is mainly due to increase in processor core count, increase in package thermal resistance, challenges in multi-chip integration and maintaining generational performance CAGR. At the same time, several other platform level components such as PCIe cards, graphics cards, SSDs and high power DIMMs are being added in the same chassis which increases the server level power density. To mitigate cooling challenges of high TDP processors, mainly two cooling technologies are deployed: Liquid cooling and advanced air cooling. To deploy liquid cooling technology for servers in data centers, huge initial capital investment is needed. Hence advanced air-cooling thermal solutions are being sought that can be used to cool higher TDP processors as well as high power non-CPU components using same server level airflow boundary conditions. Current air-cooling solutions like heat pipe heat sinks, vapor chamber heat sinks are limited by the heat transfer area, heat carrying capacity and would need significantly more area to cool higher TDP than they could handle. Passive two-phase thermosiphon (gravity dependent) heat sinks may provide intermediate level cooling between traditional air-cooled heat pipe heat sinks and liquid cooling with higher reliability, lower weight and lower cost of maintenance. This paper illustrates the experimental results of a 2U thermosiphon heat sink used in Intel reference 2U, 2 node system and compare thermal performance using traditional heat sinks solutions. The objective of this study was to showcase the increased cooling capability of the CPU by at least 20% over traditional heat sinks while maintaining cooling capability of high-power non-CPU components such as Intel’s DIMMs. This paper will also describe the methodology that will be used for DIMMs serviceability without removing CPU thermal solution, which is critical requirement from data center use perspective.

2014 ◽  
Vol 602-605 ◽  
pp. 2713-2716 ◽  
Author(s):  
Xin Rui Ding ◽  
Yu Ji Li ◽  
Zong Tao Li ◽  
Yong Tang ◽  
Bin Hai Yu ◽  
...  

LED has been regarded as the next generation lighting source. As for high power LED lamps, heat accumulation will cause a series of problems. Therefore, thermal management is very important for designing a high power LED lamp. Three types of heat sinks are designed by using the finite element analysis (FEA) method for an 180W high power LED lamp. Then the optimized heat sinks are developed and experiments are performed to demonstrate the simulated results. At the same time, the thermal performances with different working angles are investigated experimentally. The heat sink with heat pipe has a better heat dissipation performance than the conventional heat sink under the same input power. The working angles of the lamps greatly influence the thermal performance of each heat sink. For the same heat sink, the temperature varies with different install directions and working angles. Finally, the heat sink with the best thermal performance is recommended. The results have practical significance in designing high power LED lamps.


2005 ◽  
Author(s):  
John R. Hartenstine ◽  
Jerome E. Toth

Die level power densities are hitting unprecedented fluxes of over 200W/cm2, in turn driving the junction-to-case and case-to-sink resistances higher than the sink to ambient resistance. The biggest potential for thermal improvement now lies within the heat sink to device interface. Reduction in this critical interface resistance can be realized through direct attach of the heat sink base to the silicon. Challenges to this approach include the near matching of the coefficient of thermal expansion (CTE) of these two surfaces. Design concepts for a heat pipe heat sink with a CTE similar to that of silicon are investigated. Design concepts take into account materials, heat sink design configurations, and material compatibility with heat pipe working fluids. Prototype test articles were designed and fabricated using a copper/molybdenum/copper laminate as the heat input surface. Test data indicating an evaporator thermal resistance of 0.080 cm2·°C/W at 290W/cm2 are presented.


Author(s):  
Aalok Trivedi ◽  
Nikhil Lakhkar ◽  
Madhusudhan Iyengar ◽  
Michael Ellsworth ◽  
Roger Schmidt ◽  
...  

With the continuing industry trends towards smaller, faster and higher power devices, thermal management continues to be extremely important in the development of electronics. In this era of high heat fluxes, air cooling still remains the primary cooling solution in desktops mainly due to its cost. The primary goal of a good thermal design is to ensure that the chip can function at its rated frequency or speed while maintaining the junction temperature within the specified limit. The first and foremost step in measurement of thermal resistance and hence thermal characterization is accurate determination of junction temperature. Use of heat sinks as a thermal solution is well documented in the literature. Previously, the liquid cooled cold plate tester was studied using a different approach and it was concluded that the uncertainty in heat transfer coefficient was within 8% with errors in appropriate parameters, this result was supported by detailed uncertainty analysis based on Monte-Carlo simulations. However, in that study the tester was tested computationally. In this paper, testing and characterization of a heat sink tester is presented. Heat sinks were tested according to JEDEC JESD 16.1 standard for forced convection. It was observed that the error between computational and experimental values of thermal resistances was 10% for the cases considered.


Author(s):  
Matt Connors

Current blade processors need air cooling solutions that dissipate 100–300 watts with heat sinks that are less than 30 mm high. In order to cool these processors, the heat sink base has to grow in length and width to compensate for the lack of available height. As these dimensions grow, decreasing the base spreading of the heat sink becomes an important factor is decreasing the overall resistance of the heat sink. A vapor chamber used as a substitute to common copper or aluminum as the base of the heat sink can increase performance by 20–25%. A vapor chamber is a two phase heat transport system that significantly reduces the spreading resistance in applications where there is a high heat flux processor coupled with a large heat sink. In this paper, a CFD model will be constructed to predict the performance gains realized by using a vapor chamber base in lieu of a copper or aluminum base. These predictions will then be experimentally tested to confirm the modeling parameters and the actual measured thermal performance of the heat sink. By utilizing vapor chambers in heat sink design, thermal engineers will gain valuable heat sink performance within the constraints imposed by the blade system architecture.


2011 ◽  
Vol 42 (11) ◽  
pp. 1257-1262 ◽  
Author(s):  
Xiang-you Lu ◽  
Tse-Chao Hua ◽  
Yan-ping Wang

Author(s):  
Albert Chan ◽  
Jie Wei

Feasibility study on alternative cooling methods to air-cooling with heat sinks is provided in this paper. The study focuses on cooling of 64-bit microprocessor at 80nm technology node with projected heat dissipation of 200W. An example was presented to illustrate limitation of air-cooling for the 200W microprocessor using an all-Cu heat sink with tall fins. Three alternatives to air-cooling were studied in this work: liquid cooling, two-phase convective flow cooling and refrigeration cooling. Thermodynamic analysis was used to estimate operating conditions and fluid flow rates for each alternative. The information provides a preliminary basis for assessing capabilities and weaknesses among alternatives. Liquid and two-phase cooling simply transfer heat from high to low temperature. In contrast, refrigeration cooling operates as a heat pump, moving heat from low to high temperature. Refrigeration cooling offers capability to cool microprocessor (LSI) chip to temperatures below ambient or freezing. The drawback is more heat must be removed from the system. Liquid cooling operates at close to ambient pressure, while two-phase and refrigeration cooling operate at higher pressures. Challenges to implementation of all three alternatives include availability of low cost, miniature components (pumps or compressors, heat exchanger and condenser), designing for redundancy (or reliability) and ease of installation and field service. In terms of component availability and cost, liquid cooling is preferred choice, followed by two-phase and refrigeration cooling.


2013 ◽  
Vol 56 (1-2) ◽  
pp. 18-26 ◽  
Author(s):  
Ji Li ◽  
Feng Lin ◽  
Daming Wang ◽  
Wenkai Tian

2018 ◽  
Vol 225 ◽  
pp. 04009
Author(s):  
Ali Elghool ◽  
Firdaus Basrawi ◽  
Hassan Ibrahim ◽  
Thamir K Ibrahim ◽  
Shaharin A. Sulaiman ◽  
...  

Heat sink lack of design is one reason that negatively affects the performance of thermoelectric modules. As compared to conventional cooling systems equipped with thermoelectric generators (TEG), heat pipe heat sink has various points of interest. Heat pipe heat sink is the most appropriate heat exchanger for medium temperature range under 300 °C. This paper demonstrates the effect of different materials of heat pipe-heat sink on the TEG performance. Two types of heat sinks were tested with TEG, one made from copper while the other from aluminium. The aim is to improve power output of TEG by an appropriate material of fins and metal block with heat pipes. The prototype was experimentally tested and the TEG cold side temperature, voltage and current were measured in both conditions, natural and forced convection. It was found that highest power output was achieved using copper heat sink in the case of forced convection, being 7.7 W whereas, lowest power output was obtained using aluminium heat sink in the case of natural convection, being 2.67 W. It is evident that copper heat sink is more effective than aluminium heat sink in terms of power output. However, both types of heat sink needs optimisation in terms of power output, cost and economic efficiency, while the results shown in this paper are just in terms of power output.


2012 ◽  
Vol 490-495 ◽  
pp. 2278-2281
Author(s):  
Yun Jun Gou ◽  
Xiao Hui Zhong

A new cooling concept for high power LED street lamp by combining the heat release of high power LED with two-phase heat transfer heat pipes was proposed, and in this study a series of heat pipes with specific fins structure were developed. Through experimental results, we found the new heat pipe heat exchangers have the features of higher efficiency of heat dissipation and more compact construction which meets the demand of heat dissipation for high power LED than the traditional heat pipe heat exchangers and the new exchanger with outwards-radiate structure has the best heat dissipation performance.


Sign in / Sign up

Export Citation Format

Share Document