Last weekend, I was watching an online video, when my custom-built desktop machine suddenly shut down. I attempted to boot it up, and it shut down within 5 seconds. It was apparent that it was a hardware problem. The lack of POST beep pretty much confirmed this.
A lot of forum-surfing led me to believe that one or both of my RAM sticks were fried. I unseated both RAM sticks and booted the machine. There was a lot of beeping (expected due to missing RAM), but it shut down within a few seconds. RAM wasn’t the issue.
Other suggestions on the forum were related to the power supply (PSU), graphics card and CPU. The fact that I had a display meant that the nVidia 8800 GTS graphics card was just fine. I also have a 520W PSU which is more than enough. Moreover, this machine ran fine for over a year and a half.
The next time I booted the machine, I had just enough time to get into the BIOS Health Status page – CPU was at 50C and rising … and then the machine shut down. Further inspection revealed that the CPU fan had become a little loose from the motherboard. I reseated the CPU fan (after ensuring that the silica gel silicon paste (thermal paste) [thanks Amit for pointing out the error] layer was fine), and rebooted the machine. CPU temperature was under control and there are no more issues!