Steady and Transient Heat Transfer Analysis of the Microprocessor System Using One-Dimensional Thermal Network

Author(s):  
Koji Nishi ◽  
Tomoyuki Hatakeyama ◽  
Masaru Ishizuka

This paper introduces and discusses one-dimensional thermal network for microprocessor silicon die temperature prediction both under steady and transient states. Firstly, one-dimensional thermal network with average temperature nodes is constructed with the thermal spreading resistance and the thermal local resistance. Secondly, characteristics of thermal resistances under steady state are discussed. By conducting three-dimensional heat conduction simulation, thermal resistance values are evaluated for the case with different cooling solutions and different heat distributions at silicon die bottom of microprocessor. Thirdly, thermal resistance behavior of introduced thermal network during transient state is explored for further understanding of temperature transition of the microprocessor silicon die. Since the thermal spreading resistance takes relatively large value and changes its value drastically, this paper investigates transient behavior of thermal spreading resistance from TIM (Thermal Interface Material) to heat sink base as an example.

Author(s):  
Koji Nishi ◽  
Tomoyuki Hatakeyama ◽  
Shinji Nakagawa ◽  
Masaru Ishizuka

The thermal network method has a long history with thermal design of electronic equipment. In particular, a one-dimensional thermal network is useful to know the temperature and heat transfer rate along each heat transfer path. It also saves computation time and/or computation resources to obtain target temperature. However, unlike three-dimensional thermal simulation with fine pitch grids and a three-dimensional thermal network with sufficient numbers of nodes, a traditional one-dimensional thermal network cannot predict the temperature of a microprocessor silicon die hot spot with sufficient accuracy in a three-dimensional domain analysis. Therefore, this paper introduces a one-dimensional thermal network with average temperature nodes. Thermal resistance values need to be obtained to calculate target temperature in a thermal network. For this purpose, thermal resistance calculation methodology with simplified boundary conditions, which calculates thermal resistance values from an analytical solution, is also introduced in this paper. The effectiveness of the methodology is explored with a simple model of the microprocessor system. The calculated result by the methodology is compared to a three-dimensional heat conduction simulation result. It is found that the introduced technique matches the three-dimensional heat conduction simulation result well.


Author(s):  
Anjali Chauhan ◽  
Bahgat Sammakia ◽  
Kanad Ghose ◽  
Gamal Refai-Ahmed ◽  
Dereje Agonafer

The stacking of processing and memory components in a three-dimensional (3D) configuration enables the implementation of processing systems with small form factors. Such stacking shortens the interconnection length between processing and memory components to dramatically lower the memory access latencies, and contributes to significant improvements in the memory access bandwidth. Both of these factors elevate overall system performance to levels that are not realizable with prevailing and other proposed solutions. The shorter interconnection lengths in stacked architectures also enable the use of smaller drivers for the interconnections, which in turn reduces interconnection-level energy dissipations. On the down side, stacking of processing and memory components introduces a significant thermal management challenge that is rooted in the high thermal resistance of stacked designs. This paper examines and evaluates three distinct solutions that address thermal management challenges in a system that stacks DRAM components onto a processing core. We primarily focus on three different configurations of a microchannel-based single-phase liquid cooling system with a traditional air-cooled heat sink. Our evaluations, which are intended to study the limits of each solution, assume a uniform power dissipation model for the processor and accounts for the thermal resistance offered by the thermal interface material (TIM), the interconnect layer, and through-silicon vias (TSVs). The liquid-cooled microchannel heat sink shows more promising results when integrated into the package than when added to the microprocessor package from outside.


2013 ◽  
Vol 37 (3) ◽  
pp. 765-773 ◽  
Author(s):  
Farn-Shiun Hwu ◽  
Ho-Chih Cheng ◽  
Ya-Hui Hu ◽  
Gwo-Jiun Sheu

A three-dimensional numerical model using the finite element method is proposed in the present study to accurately simulate the influences of the thermal resistance on the submount of an LED. In a system with adiabatic lateral boundaries, the internal thermal resistance of the submount is principally analyzed from the series connection effect of the spreading thermal resistance and one-dimensional material resistance. However, the total thermal resistance is used for analysis under various heat dissipating conditions due to the complex coupling relations among the material resistance, the spreading thermal resistance, and the external thermal resistance. A higher contact ratio between heat source and submount, a larger external convective effect, and dissipation of heat from the symmetrical axis of the submount will decrease the spreading thermal resistance.


2016 ◽  
Vol 138 (3) ◽  
Author(s):  
Danish Ansari ◽  
Kwang-Yong Kim

The performances of various transverse-flow double-layer microchannel heat sink configurations were evaluated compared to those of parallel-flow heat sink configurations via conjugate heat transfer analysis. For the analysis, three-dimensional Navier–Stokes and energy equations for steady incompressible laminar flow were solved using a finite-volume solver. Water with temperature-dependent thermophysical properties was used as a coolant. The thermal resistances were evaluated for various flow configurations of both cross-channel and parallel-channel designs with identical geometric parameters and total flow rate. Changes in the microchannel flow direction lead to remarkable changes in thermal resistance and temperature uniformity. A transverse-flow configuration exhibited the best overall performance among the tested flow configurations in terms of the thermal resistance, temperature uniformity, and pressure drop.


Author(s):  
Peter Sterling

The synaptic connections in cat retina that link photoreceptors to ganglion cells have been analyzed quantitatively. Our approach has been to prepare serial, ultrathin sections and photograph en montage at low magnification (˜2000X) in the electron microscope. Six series, 100-300 sections long, have been prepared over the last decade. They derive from different cats but always from the same region of retina, about one degree from the center of the visual axis. The material has been analyzed by reconstructing adjacent neurons in each array and then identifying systematically the synaptic connections between arrays. Most reconstructions were done manually by tracing the outlines of processes in successive sections onto acetate sheets aligned on a cartoonist's jig. The tracings were then digitized, stacked by computer, and printed with the hidden lines removed. The results have provided rather than the usual one-dimensional account of pathways, a three-dimensional account of circuits. From this has emerged insight into the functional architecture.


2008 ◽  
Vol 67 (1) ◽  
pp. 51-60 ◽  
Author(s):  
Stefano Passini

The relation between authoritarianism and social dominance orientation was analyzed, with authoritarianism measured using a three-dimensional scale. The implicit multidimensional structure (authoritarian submission, conventionalism, authoritarian aggression) of Altemeyer’s (1981, 1988) conceptualization of authoritarianism is inconsistent with its one-dimensional methodological operationalization. The dimensionality of authoritarianism was investigated using confirmatory factor analysis in a sample of 713 university students. As hypothesized, the three-factor model fit the data significantly better than the one-factor model. Regression analyses revealed that only authoritarian aggression was related to social dominance orientation. That is, only intolerance of deviance was related to high social dominance, whereas submissiveness was not.


Sign in / Sign up

Export Citation Format

Share Document