简介：某天，电脑开始随机重启。没有发现任何规律。每天发生 1 到 8-9 次，通常为 2-5 次。两次发生间隔为 2-3 天。重启不会导致 BSOD。事件日志没有提供任何提示。温度不高 - 发生在空闲负载期间。详情如下，并在我的第一个问题中。
目前主板刚刚换了技嘉的，CPU也换了，RAM也换了，但是随机重启的问题还是存在。
下面是对发生重启的排列的粗略图形指南。
钥匙：
WP = Main or Wall Power source
IP = Intermediate Power Source a.k.a UPS
PSU = Computer Power Supply Unit - Corsair 550M
Cs = Computer case/cabinet
MB = Motherboard GA-H97-D3H
RAM - RC = 2x Corsair 4GB DDR3 1333
RK1 = 1x Kingston 8GB DDR3 1600
RK2 = 2x Kingston 8GB DDR3 1866
CPU = i5-4440
HDDs = WD 1TB Win 7 + Seagate 2TB OpenSuSE Linux Tumbleweed
OS = within Win7 / Linux / UEFI
No add-on graphics ever (at my end), only iGPU.
这些是系统测试过的场景。启动时 FAIL 表示重现重启；PASS 表示未看到重启。
如果有后缀 - 则表示组件已更改。尾随?表示该组件可能不存在。
按年代顺序，
FAIL: WP - IP - PSU - Cs - MB - RC - CPU - HDDs - Win
FAIL: WP - IP - PSU - Cs - MB - RC - CPU - HDDs - Linux
FAIL: WP - IP - PSU - Cs - MB - RC - CPU - HDDs - BIOS
--- comp sent to dealer without HDDs --->
FAIL: WP_dlr - IP_dlr? - PSU - Cs - MB - RC - CPU - HDDs_dlr - Win_dlr
PASS: WP_dlr - IP_dlr? - PSU_dlr - Cs_dlr - MB - RC - CPU - HDDs_dlr - Win_dlr
(^ tested for less than a day)
--- comp returned to me with dealer PSU and Cs --->
FAIL: WP - IP - PSU_dlr - Cs_dlr - MB - RC - CPU - HDDs - Win
FAIL: WP - IP - PSU_dlr - Cs_dlr - MB - RK1 - CPU - HDDs - Win
(^ a week passed without restarts after switching to ram RK1)
FAIL: WP - IP - PSU_dlr - Cs_dlr - MB - RK1 - CPU - HDDs - Linux
继续。
--- comp returned to dealer without HDDs --->
--- MB only forwarded to Gigabyte --->
PASS: WP_gb - IP_gb? - PSU_gb - Cs_gb? - MB - R_gb - CPU_gb - HDDs_gb? - ?
(^ burn-in test for 2 days)
FAIL: WP_gb - IP_gb? - PSU_gb - Cs_gb? - MB - RC - CPU - HDDs_gb? - ?
(^ sent my CPU + Corsair RAM)
--- Gigabyte return items to dealer; say one of the RC sticks to blame --->
PASS: WP_dlr - IP_dlr? - PSU - Cs - MB - 1xRC + RK1 - CPU - HDDs_dlr - Win_dlr
(^ tested intermittently over 3-4 days)
--- comp returned to me with "faulty" RAM removed --->
FAIL: WP - IP - PSU - Cs - MB - 1xRC + RK1 - CPU - HDDs - Win
--- MB+CPU+RAM(RC+RK1) resent to Gigabyte --->
(no testing done; CPU examined - observed some pins have blackened; say CPU to blame)
--- dealer replaces CPU with same model --->
PASS: WP_dlr - IP_dlr? - PSU_dlr - Cs_dlr - MB - 1xRC + RK1 - CPU_new - HDDs_dlr - Win_dlr
(^ tested for 3 days)
继续。
--- comp returned to me with new CPU --->
FAIL: WP - IP - PSU - Cs - MB - RC + RK1 - CPU_new - HDDs - Win
FAIL: WP - IP - PSU - Cs - MB - RC + RK1 - CPU_new - HDDs - BIOS
--- MB+CPU_new+RAM(RC+RK1) sent to Gigabyte --->
PASS: WP_gb - IP_gb? - PSU_gb - Cs_gb? - MB - RC+RK1 - CPU_new - HDDs_gb? - ?
(^ testing setup and duration unknown but reported after 3 days; say PSU to blame)
At this stage, I call Gigabyte and explain testing history to site manager,
who offers replacement MB.
--- comp returned to me with RMAed MB --->
FAIL: WP - IP - PSU - Cs - MB_rma - RC + RK1 - CPU_new - HDDs - Win
PASS: WP - IP - PSU - Cs - MB_rma - RC + RK1 - CPU_new - HDDs - Linux
(^ a few hours)
FAIL: WP - IP - PSU - Cs - MB_rma - RC + RK1 - CPU_new - HDDs - BIOS
--- swap RAM with another (working) comp, yesterday --->
FAIL: WP - IP - PSU - Cs - MB_rma - RK2 - CPU_new - HDDs - Win
(PASS: WP_oth - PSU_oth - Cs_oth - MB_oth - RC + RK1 - CPU_oth - HDD_oth - Win)
现在考虑我的下一步测试步骤 可能是将其升级到神学部门，但会再检查一下时间建议。
如果您阅读了最初的问题，您会注意到原始 MB 已于 6 月发送给技嘉。它被送回维修处，并且其 BIOS 升级到版本 F7。之前的版本是 F3。新版本增加了对第五代处理器的支持，它们还可以防止 BIOS 回滚。新 BIOS 可能是罪魁祸首吗？ RMAed 主板也有 F7。与这种可能性相反的是，升级发生在 6 月，重启行为始于 10 月中旬。
自从技嘉第一次将硬盘返还给我后，硬盘消失的问题就再也没有出现过。在 Windows 或 Linux 的任何使用过程中都没有出现崩溃。
我可能会将 PSU 与正常工作的计算机进行交换，尽管早期的配置已使用经销商的 PSU 进行了测试，但没有成功。
谢谢阅读。有什么想法吗？