基于飞腾处理器的瘦客户机散热结构设计

计算机工程与科学

基于飞腾处理器的瘦客户机散热结构设计

李红,李俊,龚国辉

（湖南长城银河科技有限公司,湖南长沙 410205）

收稿日期:2017-12-01 修回日期:2018-03-22 出版日期:2018-10-25 发布日期:2018-10-25

Heat dispersion structure design of thin

client computers based on FT processor

LI Hong,LI Jun,GONG Guohui

（Hunan Great Wall Galaxy Science and Technology Ltd.,Changsha 410205,China）

Received:2017-12-01 Revised:2018-03-22 Online:2018-10-25 Published:2018-10-25

摘要/Abstract

摘要：

基于目前国产处理器的瘦客户机整机功耗相对较高,利用常用的无风扇散热的被动散热方式在整机散热性能方面稍显不足,尤其是存在散热死区,从而设计了一种基于飞腾四核处理器的瘦客户机。该瘦客户机采用双散热模组设计,采用4个进风道和2个出风道设计,使得机体内部无散热死区,实现整机超强的散热性能。温度测试数据表明：在25℃环境中,机器满负荷工况运行时,机体内各测试点的温度均在52℃以内,满足所有元器件的工作温度要求；同时,机器箱体外表面温度低于36℃,用户体验感良好。

关键词: 瘦客户机, 飞腾处理器, 散热结构, 散热模组, 风道

Abstract:

Due to the relatively high power consumption of thin client computers based on current domestic processors, the commonly used passive heat dispersion method with the fanfree thermal conductivity is slightly inadequate in the aspect of heat dispersion performance of the whole machine, having the dead zone of heat dispersion in particular. We design a thin client computer based on quadcore FT processor. The thin client computer has a double thermal module, 4 inlet ducts and 2 exhaust ducts, so that there is no heat dead zone inside of the client computer, and it can achieve a good heat dispersion performance of the whole machine. Temperature test data show that, when the computer runs under full load condition in the environment of 25℃the temperature of each test point in the body is within 52℃ , which meets the working temperature requirements of all components. Besides, the outer surface temperature of the machine is lower than 36℃, which brings favorable user experience.

Key words: thin client computer, FT processor, heat dispersion structure, thermal module, air passage

李红,李俊,龚国辉. 基于飞腾处理器的瘦客户机散热结构设计[J]. 计算机工程与科学.

LI Hong,LI Jun,GONG Guohui.

Heat dispersion structure design of thin

client computers based on FT processor

[J]. Computer Engineering & Science.

[1]	李红，李俊，龚国辉. 笔记本电脑散热结构优化设计[J]. 计算机工程与科学, 2019, 41(03): 446-451.
[2]	方建滨，杜琦，唐滔，陈顼颢，黄春，杨灿群. 飞腾处理器与商用处理器性能比较[J]. 计算机工程与科学, 2019, 41(01): 1-8.
[3]	毛席龙，杨安，吕高锋，林琦，程辉. 基于可变步长的访存延迟测量模型的研究与实现[J]. J4, 2014, 36(01): 12-18.