Bad Ram?

[Tachyon74]

So I have a new compy,

Intel e8400
2x 1gb crucial ram
150gb WD raptor
Gigabyte p35-ds4
evga 8800gt 512mb

got adventurous and put Vista 64 on it. No major problems until just a few days ago.
Suddenly I have a lot of bluescreens and random restarts, and what's more the thing seems to crash while it's starting up sometimes. I booted into XP thinking that Vista finally just killed itself, and it got a bluescreen after a while too. My CPU seems like it's staying at a good temp, always right around 33C (stock heatsink, so not fantastic, but it seems to be working).

So I ran memtest and got a ton of errors. I pulled one bank of ram and the errors go away. Problem solved right? Except then I started getting bluescreens and crashes again. I let memtest run like 16 passes on the other bank and it doesn't fail, and it still does on the one.

Am I looking at a bad mobo or did both of my ram sticks die and memtest just isn't showing it? I ordered some more ram to try out, it will be here on tuesday, but until then, any suggestions?

 

[jackschmittusa]

Bad ram can corrupt files in Windows so that the os is still buggy after removing/replacing the defective ram.

 

[robisbell]

okay, make and model of Motherboard, and ram. bios revision of motherboard.

 

[Tachyon74]

Mobo is the gigabyte p35-ds4 CPUZ reads it as rev A2 so I think that means it is the 2.0 rather than the 2.1.

*Edit* Also BIOS Version is F11 (released 1/2/08)
I realized that would be useful


Memory is Crucial Ballistix BL12864aa1065.8fe5

Both sticks cause a lot of errors in memtest86 although now I have noticed that either one or the other doesn't seem to error a lot on their own, because I just swapped them and ran it again.

 

[robisbell]

all of the ram are the exact same model number?

 

[robisbell]

http://www.crucial.com/store/l....aspx?model=GA-P35-DS4

I'd suggest you get "The Ultimate Boot CD" and have it run memtest86+ on each stick individually for 6 hours each, and report back the results from both tests. I know you have run memtest basic, but I'd like this run and to see what errors it's throwing.
it may be bad ram, or may be a defective mobo, want to clarify something 1st.

 

[cubby1223]

Crucial Ballistix ram demands a higher than normal voltage, default is 1.8v, your ram demands 2.2v. Do you have this set in the bios?

 

[Tachyon74]

Actually, I didn't have the voltage set very carefully, which is something I looked at last night. I went into BIOS to underclock the RAM to DDR-800 speed (its rated at ddr 1066) and it seemed to run more stably. When I set it back to default it threw me a warning related to the voltage levels and I changed a setting that let the BIOS manage the voltage levels. I ran memtest all night (went and downloaded the new version of memtest86) without errors. I will double check the level its at when I get home, but I think I might have accidentally set the voltage to the wrong level when I was changing some BIOS settings a few days ago.

@robisbell: Yes, both sticks are identical, I bought a 2gb "kit". I will look into memtest86+ if I have any more trouble though

 

[Tachyon74]

Ok so how do I narrow down the problem to being a motherboard problem? I have tried both clocking the ram down to 800mhz or giving it 2.2v and clocking it back to 1066 (what it should be rated for) and I have still had problems either way. The reason I ask about the motherboard is on one reboot my BIOS complained that it failed checksum and I had to re-flash it. Then I had another bluescreen and it rebooted and when I tried to open the boot-time flash utility, I got a screen full of scrambled ascii characters (it worked just fine before).
And yet it seems to work ok some of the time (I'm writing on it right now...)

*EDIT*
So I downloaded the newest memtest86+ and burned a cd from it. I am getting a ton of failures right off the bat on test 2 on both sticks, one stick, AND the other stick.
Just for kicks, I tried changing the speed down to DDR-800 again but no help.
So assuming memtest86+ is more robust, then it definitely looks like bad RAM. Not sure what the deal is with my BIOS acting up, but my replacement RAM shows up tomorrow, so we'll see what happens then

 

[Tachyon74]

Update!
So I got the new RAM today. Put it in... and still had memtest failures. So I tried to start from scratch... I re-flashed the BIOS to the latest version, and ran a load optimized defaults. I noticed this stuck voltage control on manual, although all the sub voltages were set to auto. Then I noticed something... the CPU core voltage was set "manually" to 1.25v and I remembered that the last time I looked at CPU-z, my core voltage went between 1.1 and 1.8 depending on if there was a load or not. I am running memtest again and no errors so far. I'm going to let it keep running for a few hours. So I have a question: can giving the CPU too low of a voltage be causing trouble?
I don't get it because the memory is running a little under with the default settings, but the CPU seems like it runs under when I went and set the memory voltage individually with everything else on auto. Also there's that issue with the BIOS checksum and the bios built in flash util?

any more advice or stuff to try?