Random restarts with Eclipse SLI board

[conruse]

This is a copy from my post on the Vista forums havent gotten much help there.

Describe the problem in detail!

I just built a new system on December 22nd 2008. I do Tier 2 and 3 Helpdesk work so I can work my way around a system. I put the machine together installed Vista did all the updates got all the newest drivers. Basically since I have built the Machine it will just randomly restart. It will not display a BSOD, I have auto restart turned off, and it does not create an event in the event viewer. It is almost like someone touched the reset button.

When did the problem start?

December 22nd 2008 the day I built the machine.

What have you tried to fix it?

Tried several different BIOS setups. I have upgraed the bios, did all the windows updates, got all the newest drivers for video, audio, lan etc etc. Removed MSI Live Monitor, disabled green power in Bios, and removed overclocking center. Went through event viewer and corrected three of the four errors that was showing up. One error is the Kerebos one and I am not worried about that one.  I have just run both Memtest86 and Memtest86+ and both of them stop responding about half way through the test.

What is the error message (of any)?

I have Vista set to not auto restart so I should see a BSOD but I get no message. I have had 4 restarts just today and none of them created a new MEMORY.DMP file or any type of event in the viewer.

I do get warnings about something called FLASH.SYS but I cannot seem to track down what this is.

What was the last thing you did before the problem started?

Built the machine.

What are your system specs?

Vista 64
MSI Eclipse SLI X58
12 Gigs Crucial ram matched to the MoBo using the crucial matching system on the crucial website. The MoBo supports up to 24 and I am using Vista 64bit.
Antec P190 Case (Dual power supplies for a total of 1.2 KW of power)
Intel Core I7 2.66
Creative Extreme Gamer Edition PCI E sound card
Mcaffee
BFG 260 GTX
Coolmaster V8 cooler (CPU under stress runs about 70f well below the max of 208f)


Do you have a screenshot?

Nope


My next step is to take out 6 gigs of RAM and see if the issue goes away.

Any suggestions?

 

[Jack the Newbie]

12 Gigs Crucial ram matched to the MoBo using the crucial matching system on the crucial website.

Please provide a link to the Crucial Product site for your memory modules and make a screenshot of the CPU-Z "SPD" view.

My next step is to take out 6 gigs of RAM and see if the issue goes away.

Test one memory module only and see what happens.  If things seem fine, add one more module and test again (and so on).

 

[conruse]

Here is the link to the acutal ram I bought from New Egg

http://www.newegg.com/Product/Product.aspx?Item=N82E16820148221

Here is a link to what Crucial recommended.

http://www.crucial.com/store/listparts.aspx?model=7520-010%20%28Eclipse%20SLI%29

Here is the SPD View.  The only thing that changes in all 6 slots is the serial number.  Otherwise all the numbers match up.

 

[conruse]

Scratch this part.  I just had it reset after running stable for a few hours.  I was playing WoW and chatting on vent.

 

[conruse]

I just had 3 reboots within a 30 minute span so I pulled three sticks and left the other three in based on the MoBo ram config.  So far it seems stable I am going to run mem86 and see if it can clear the memory.

 

[HU16E]

Just as a reminder, with all the memory swapping, to keep the population for the board correct. 1 stick in slot #2 (AO), 2 sticks slot #2 (AO) & #4 (BO), 3 sticks slot #2 (AO), slot #4 (BO), & slot #6 (CO), etc. as per the manuals population rules.

 

[conruse]

HU16e Thanks for the reminder.  I did pull out the manual to make sure I was putting the three 2 gig sticks in the right slot.

So far it passed both versions of Memory Test 86 and right now I am burning in Prime 95 on torture/blend. 

So far it looks like the RAM was the issue.  Maybe having all six dimms filled was just too much for the board.  I know back in the day it was recommended to never fill all the dimms but I thought newer boards did not have that issue.

 

[conruse]

Two days now of running rock solid it passed all the mem 86 tests and ran Prime 95 torture/blend mode without any issues.  At this point only time will tell if going down to 6 gigs fixed it.

I guess my question is why does MSI build a board that can handle up to 24 gigs but when you put 12 in a 64 bit system it gets unstable?

I tested all my RAM and they all passed.  I can only hope the vendor will take the RAM back.

 

[HU16E]

That's great news on getting your system stable. The MSI question about support for 24G is a question many of us ask about a lot of manufacturers, i.e. ddr2 mem. marked as 1066 & only 800 chipped, ddr3 1333 & only 1066, etc. But, this is the wrong section of the forum for us to mount a frustration campaign. Again, glad to hear your system is up & runnin'.