Jump to content
Corsair Community

2 pairs of TWIN2X4096-8500C5D BSOD after 12+hrs prime95


markmark

Recommended Posts

I have 8gb ram, with two sets of pc8500 dominator ram. I have run each of the 4 sticks, one at a time, through all the dimm slots and memtest+. Everything passes, including all 4 dimms in place for 14 hrs of memtest+

 

I have DRAM voltage at 2.1v

NB voltage at 1.4v - 1.44v

DIMM speed at the lowest value above "auto"

CAS settings 5-5-5-15

cpuv 1.35

antec 900 case with three 120mm inlet fans, one 120mm outlet fan, one 200mm top outlet fan.

Q9400 cpu

 

I have tried running at 333FSB to 450 (which changes the DIMM Speed accordingling, I have the "strap" at 333). My latest try was with FSB at 440.

 

After running prime95 on all 4 cpu's for about 1/2 day or so, I repeatedly get a BSOD. It doesn't seem to matter what FSB speed I am using.

 

Once I get the BSOD, if I boot back into memtest+, it will show errors. If I shut it down again and wait a few minutes (cooling down???), it will boot into memtest+ with no errors for many hours.

 

I have had to reinstall the os several times due to all the bsod's and the os getting corrupted. After enough bsod's, the os refuses to boot.

 

Note, CPU-Z reports the memory chips as PC2-6400 (400 MHz) Max Bandwidth, with part number CM2x2048-8500C5D. Why would it say that rather than PC2-8500 (533 MHz) CM2x4096-8500C5D?

 

During the last round this afternoon, i had speedfan running monitoring my cpu temps, and they did not spike, and all of the temps stayed between 48C and 55C the entire time.

 

I bought 1066 ram so I would have additional headroom over 800 ram. But I keep seeing in this forum i need to stay at 800. why is this??? when I paid for 1066?

 

what can i do to get this system working?

Link to comment
Share on other sites

If you pass memtest one up and all four modules populating, then you are stable with the memory. The issue is external to the memory.

 

You have an X38 chipset and they have a default MCH of 1.1v. You should not exceed 1.3v with that chipset. I usually advise +.1v --> 1.2v and 1.25v at a max.

Link to comment
Share on other sites

I bought 1066 ram so I would have additional headroom over 800 ram. But I keep seeing in this forum i need to stay at 800. why is this??? when I paid for 1066?

 

The memory is SPD (Serial Presence Detect) enabled and programmed to JEDEC Standard values of 6400C5 @ 1.8v. This is to make it plug and play compatible. However, if you want to access the greater than JEDEC memory speeds/timings, you need to manually enter the necessary values.

 

For example, in the CPU-z SPD tabs the Max bandwidth is JEDEC DDR3 1333Mhz @ 9-9-9-24 2T. The part number shows my actual modules designation of CM3X2G1866C9D which is 1866Mhz 9-9-9-24 2T.

 

Once I get the BSOD, if I boot back into memtest+, it will show errors. If I shut it down again and wait a few minutes (cooling down???), it will boot into memtest+ with no errors for many hours.

 

You speak of a BSOD which immediately upon memtest shows errors. This leads me to think that the modules are either getting too hot or they are being overvolted. You speak of testing the temp of the CPU but you need to be certain that applying many hours of Prime and Memtesting does not damage components due to over heating/volting. The example you give is a common thermal one.

 

Also, do not be certain that your memory controller will run all four modules at the full speed of two modules. If the memory passes, and the CPU passes then your issue is the memory controller will not run all four modules at the full speed of a two module set.

 

Isolate the issue logically. The main variables to this issue are Heat <--> CPU <--> Memory Controller <--> Memory. You have already Memtested.

Link to comment
Share on other sites

Derek, I assume by setting the memory settings at 2.1v 881 and 5-5-5-15 (2T is a default and I don't see anywhere to change it), I am not stressing the memory to the full 1066 speed it is rated at. But with 2.1v, and 5-5-5-15 I am correctly configured to run it at 1066... is that what you are saying?

 

What "changes" when I change from say running ddr2-800 to ddr2-1066? Isn't the bulk of the speed of the computer determined by the cpu, as long as the ram can keep up with the FSB speed?

 

your comments lead me to think YOU think this is a memory controller issue.

 

Where is the memory controller? on the motherboard or in the cpu?

 

Is there some kind of test I can do on the memory controller to isolate/test it??

 

Why would the hardware have issues with running 100% for extended periods of time? I use pizza box computers at work that run 100% for weeks at a time with large engineering jobs. The cpu temps are nowhere near the temp they will automatically throttle back. I have LOTS of air moving in my case.

 

thanks

Link to comment
Share on other sites

  • Corsair Employees
With 4 modules I would suggest setting the memory frequency at DDR667 or DDR667 and set the memory Voltage to 2.1 Volts and set the NB/MCH/SPP Voltage to +.2 Volts as well and test the system with http://www.memtest.org. In addition, with some MB's (Mostly ASUS) you have to disable legacy USB in the bios when running any memory test.
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...