论文部分内容阅读
对于大、中型计算机系统的热设计,以往设计人员通常采用最劣法,即满足置换系统中最坏情况所对应的发热量,以及取定最大功率组件的功耗或结温(T_(Jmax))为参考依据,使得其值不超过某一指定值。显然,最劣法所描绘的热场,其反映出最坏情况下的极限参数。 而在实际的整机系统中,成千上万个电路组件的功耗是有很大差异的,其对应的结温差别也是很大的。然而,最大功耗组件和最小功耗组件在系统中所占的比例数也是小的。因此,若用这两种极端的功耗组件作为设计依据,去说明或描绘系统的功耗、结温和系统的温度分布(即:最劣法),最终得到一个与实际符合程度很小的结果。
For the thermal design of large and medium-sized computer systems, in the past, designers often adopted the worst-case method to meet the worst-case heat generation in the replacement system and determine the maximum power consumption or junction temperature (T Jmax) ) As a reference, so that its value does not exceed a specified value. Obviously, the worst law describes the thermal field, which reflects the worst-case limit parameters. In the actual whole system, the power consumption of thousands of circuit components is very different, and the corresponding junction temperature difference is also great. However, the maximum power consumption components and the minimum power consumption components in the system, the proportion of the number is small. Therefore, using these two extreme power components as a design basis to illustrate or characterize the system’s power consumption, junction temperature, and system temperature distribution (ie, the worst-case approach) will ultimately result in a less-than-realistic result .