I have a watercooled i7 3770k sitting in an Asus Maximus V Formula (that is part of the same loop using the built in VRM cooling). The CPU has always been stock and the only overclock of any kind is the XMP profile on my Kingston Hyper-X ram to set it at the rated DDR3-2133. Last week I discovered that my pump (an MCP-355) had died. I do not know for how long it had been dead. The system is left on 24/7 so the CPU may have been sitting at a very high temperature for an extended amount of time.
I have 16GB of Kingston Hyper-X ram (4x4GB) and 2x EVGA GTX 670 FTW GPUs in this system.
I finally got the replacement pump installed (an MCP35x with heatsink to keep it from burning out like the last one) and started the system back up. I had assumed that the CPU has built in thermal protection so should be fine however I am seeing very strange behavior now out of the computer. With all 4 sticks of ram installed I get a 55 post code which is "memory not installed". I get the same code with 2 sticks installed if I set the XMP profile and do not leave it at auto. I tried both sets of sticks in both sets of DIMMs on the motherboard. Furthermore, I am seeing only one of my GPUs. Both in the BIOS and the OS even though both are installed. The one GPU that is showing up is only running at PCIex8 rather than the expected x16 if there is only one GPU installed. I will try some combinations of GPUs in various slots but currently it is clearly not working as expected.
Has anyone else seen any CPU failure like this where it appears the PCIe controller and the Memory controller may be damaged simply from extended periods of high temperature? As far as I can tell the system never rebooted itself due to heat but it was definitely hot.
Any help or suggestions for troubleshooting is appreciated - thanks!
Bookmarks