MSI P55 GD85 - BSOD - MEMORY_MANAGEMENT

[wlamers]

Dear Ollem,

I am facing the exact same problem. This is my hardware config:

- MSI P55 GD-85 (BIOS 1.3)
- Intel core i7-875k
- Geil black dragin quad channel 8gb 4x2gb GB38GB1333C7QC
- Windows 7 64bit

This looks similar to your configuration. I have also problems during the first minutes after a boot in the morning (cold system). I keep getting MEMORY_MENAGEMENT BSOD, and the event log shows kernel_power error 41.

Memtest shows errors in banks 3 and 4, which disappear after several reboots. I removed the modules in bank 3 and 4. THis solved the problem for a week (I keep checking with memtest every morning). But this morning I also got errors on bank 2. I already swapt modules from banks 3-4 to 1-2 and vice versa. The problems kept appearing on bank 3-4 which pointed to faulty memory banks or cpu errors and not the memory modules itself. But now I am confused since bank 2 also shows errors.

When the system is warmed up (after several reboots) is runs stable. Intel burn test is no problem at all. Temeprature of the CPU does not go beyond 58 deg C. The power supply is also more than suffient for the system.

I also played with the CPU setting and the memory voltage which does not help at all.

I hope you have found a solution in the mainwhile. Or does anybody have any clues?


Best regards,
W.Lamers

 

[Jack the Newbie]

Please take screenshots of the CPU-Z "SPD" and "Memory" view, upload the images somewhere and include the links here for comparison.

MSI P55 GD-85 (BIOS 1.3)
- Intel core i7-875k
- Geil black dragin quad channel 8gb 4x2gb GB38GB1333C7QC
- Windows 7 64bit

What video card and PSU are you using?

 

[claykin]

For the moment I'll assume your PSU is capable (but please add the PSU, video card and other peripherals connected to PCI slots to your forum config).  I'll also assume you set the BIOS defaults and it still causes problems?

If the above are true, it sure sounds like those Geil RAM modules are not compatible.  Doubt you have 2 or 3 bad modules.  That Geil model has a CAS latency of 7 and a factory voltage setting of 1.5V.  The JEDEC spec for DDR3-1333 is CAS latency 9.  Most overclocked modules that use a CAS 7 are also using 1.65V.  Maybe try setting the timings manually in BIOS to 9-9-9 and leave 1.5V.

The issue I have with companies like Geil is that they have the chip package OEM'd for them with the "Geil" name on it.  You have no clue whose die they are using.  Its probably UMC or someone second tier and likely not Micron, Samsung, etc..  Plus they take creative license in programming the SPD chips on the modules and the end result is often what you are experiencing today.

If you cannot get them to run reliably with the CAS 9 setting and 1.5V, you can try 1.65V but be very careful overvolting RAM.  Bad things can happen.

Otherwise, do yourself a favor and return those modules and exchange them for PC1333 Crucial (Micron) modules.  You should be a happy camper after that.

 

[HU16E]

claykin is right about those Crucials. I just did some quick testing with one stick. This was with only 1.50V's for both speeds.

 

[Henry]

I've also had mine up to 1600 on 1.5V. Don't need it that high but does it no problem. The Crucial 1333 RAM is decent price does what it's supposed to do without complaining and just because it's plain black chip on a green stick ugly far too many people refuse to buy it.

 

[wlamers]

Thank your your replies!

Here I have the screenshots from cpu-z:






My bios settings are standard, but I also expirimented with those. I tried setting all CPU 'energy' settings to disabled. Runing the memory on 1.6 and 1.65 volts does not make any difference. The strange thing is that when I reallocated the memory modules (hussle them) I still have memory errors on the same addresses in bank 3-4. This suggests maybe that the problem is not the memory modules.

The full spec of the pc is:
Processor - 1 x Intel Core i7 875K / 2.93 GHz - LGA1156 (running at standard multiplier and base clock)
MSI P55-GD85 (bios version 1.3)
EnerMax Pro82+ II  425 Watt
GeIL Black Dragon Quad Channel Kit  8 GB 4 x 2 GB  DDR3 - 1333 MHz / PC3-10660 CL7 - 1.5 V GB38GB1333C7QC
Scythe Mugen 2 rev.B CPU cooler
ATI FirePro V4800 GPU
Plextor PX-880SA  DVD?RW
Crucial C300 64gb SSD
Running on Windows 7 64bit (irrelevant since MemTest shows errors prior to windows boot)

Tommorow I can try setting the timings to 9-9-9. Today I can not do this since the system is warm and running stable at the momemt. I also tried relocating the CPU, cooler and memory modules but this does not make a difference. I hope you guys have another clue.


Thank your your support and best regards,

W.Lamers

 

[Jack the Newbie]

Reduce the memory clock speed to DDR3-1066 for testing.

 

[wlamers]

I finally found a way to overcome the problem. When I start the computer and do a MemTest I still have a lot of errors. When I then shut down the computer and remove the power cable for about 10 seconds and restart the computer the errors are totally gone and the system runs stable. Anyone an idea how this is related to a defective hardware component (e.g. MB, CPU, PSU, memory, ...)?

The supplier did ran checks on the system (RMA) but did not find a problem (probably due to the time the computer was of the power grid during shipping). The now send me other memeory modules which may solve the problem.