Need Serious Help.. Can't figure out whats wrong! my CPU??

So I have a few threads currently going, one here in CPUs, somein WCing, and some in Discussion. Some of them have kind of segued into trying to solve my pc issue, so I figured its best to consolidate the discussion in a new thread because I am UTTERLY LOST!

So in my sig is what I have been running for like 2.5 years. I ran it almost always at an undervolted (except +.2 on gmch) and 8x400 with my ram in 1:1. About 3 months ago I had some odd freezing during post. This is on my GA-965P-S3 motherboard. I finally go it to stop and immediately reduced speeds to stock. Since then, the system was running fairly fine, except for periodic bsods.

I finally found sometime and money to fix the problem; which I attributed to either failing ram or a failing mobo. The S3 wasn't really built for long term OCs I feel.

So i figured, since I wanted to upgrade my board anyways, to start by replacing my motherboard and then check the ram.

I upgraded to the GA-EP45-UD3R rev 1.1 board. I swapped all the parts over, including my enermax 630W PSU, which is already an RMA'd psu, and everything ran fine for about 3 to 4 hrs. Windows 7 installed perfectly fine and I was able to watch niptuck on my tuner. Then, it began freezing in OZ and the monitor kicking off.

Then, it began freezing in the post and the bios, if it got that far. Then, the monitor became unresponsive.

So, I began swapping through each of my four sticks of OCZ ram, of which I am 100% confident 1 is functional (so I know it should work), and the same issue persisted. It eventually got to the point where the monitor would even kick on. I also tried resetting cmos multiple times. As well, I tried my second PSU, an ultra X3 600w, and still no fix.

So, to rule out the motherboard, I swapped back to my 965P board and on first boot, the monitor kicked on, but during the post screens, there were tons of odd characters surrounding the normal ones. then, within a few seconds, the monitor kicked off, never to come back on again.

So no, I just returned to my system and removed the 965Ps cmos battery for about 10 minutes. Now, the system booted up and loaded right into memtest, however, while typing, at about 7mins into the pass (like 60%), the monitor shut off. Just before this breakthrough, I was so certain it was my CPU, not I have no idea..

So I feel like this could only be one of two things.

My VGA, which I highly doubt could be doing this.
My CPU


I feel like it cannot be the ram after my extensive testing described above.
I know its not the PSUs since I have two and the problem still persists.
I feel like it cannot be the Mobos since it is happening with both..

And note, I never had to fine tune settings on my 965P to function, besides up ram voltages to 2.1V.


I am so lost, just when I start to think i know what the issue is, something different happens and throws me off. Has anyone experienced this.

Please help me!

Also, note that I have seen odd symbols in the post on both motherboards.. What cause this? I feel liek it could only honestly be my VGA or my CPU. I just feel liek a deffective vga wouldnt behave this way.. Oh and I rebooted and now I am at 8 minutes in memtest on the old board.

I am on the phone with EVGA right now to see what they think.

And it is definitely not the monitor, it still works fine with my laptop and xbox.

one pass of memtest down on the old mobo with one stick of ram.. this is really weird..


UPDATE: I got about 45mins of memtest. I then decided to hook up a HDD and install XP. I am going to try to run a CPU and a VGA stress test and see if I can figure out who is the culprit..

Guy at EVGA told me something interesting. That static electricity damage in hardware is incredibly unpredictable and that sometimes it can take months to actually manifest. hmmm


UPDATE:

In XP right now, OCCT is running cpu test. At two minutes and 34degrees on both cores..

phil178821 said:

Also, note that I have seen odd symbols in the post on both motherboards.. What cause this? I feel liek it could only honestly be my VGA or my CPU. I just feel liek a deffective vga wouldnt behave this way.. Oh and I rebooted and now I am at 8 minutes in memtest on the old board.

I am on the phone with EVGA right now to see what they think.

And it is definitely not the monitor, it still works fine with my laptop and xbox.

one pass of memtest down on the old mobo with one stick of ram.. this is really weird..


UPDATE: I got about 45mins of memtest. I then decided to hook up a HDD and install XP. I am going to try to run a CPU and a VGA stress test and see if I can figure out who is the culprit..

Guy at EVGA told me something interesting. That static electricity damage in hardware is incredibly unpredictable and that sometimes it can take months to actually manifest. hmmm


UPDATE:

In XP right now, OCCT is running cpu test. At two minutes and 34degrees on both cores..

Click to expand...

UPDATE:

Past OCCT for about 30 mins. Although that is really not long enough, I figure that if my cpu is damaged to the point to cause what was happening, then it must fail OCCT pretty fast.

So it looks like its my VGA. It starts artifact - ing 2 seconds into ATI TOOL scan.

try using the integrated graphics on the mobo and running prime95. also, try using any old s775 cpu laying around from an old build if you have one.

Well I feel like it has to be the Graphics card. I bet I fried it with careless touching.

I would think something as bad as a damaged cpu would error really fast.
From what I ahve heard, artifact-ing in bios and post is generally cpu, ram, or gpu. I ruled out ram, CPU survived 30 mins of OCCT and GPU FAILED Immediately.

Now on a sad notes, my spare cpus and graphics cards are at home.. really smart, huh.

and, my two boards do not have onboard vid support.

with all the evidence listed above, do any of you still doubt its my graphics card? I am ready, after days of test, to settle on that conclusion, unless some of you still have doubts.

Thanks a lot guys

phil178821 said:

UPDATE:

Past OCCT for about 30 mins. Although that is really not long enough, I figure that if my cpu is damaged to the point to cause what was happening, then it must fail OCCT pretty fast.

So it looks like its my VGA. It starts artifact - ing 2 seconds into ATI TOOL scan.

Click to expand...

Do you have another GPU you could try? It always helps to keep an extra one around just for this kind of problem. Also, an extra CPU and extra known good sticks of RAM are things I like to keep around just to check in case something goes crazy with my rig. The best I have for a CPU is an old 775 P4 in my wifes rig, but I also have an extra set of DDR2 sticks. Sometimes having a local friend that will let you borrow parts can be helpful as well.

Sounds like the video card.I always keeep a spare video card.LIke right now my spare is a PCI FX5200.This way it will work with any system and it was cheap.And i run SLI so i almost always have a extra pci-e video card

Artifacts like that are almost surely something video related.

Z

I do have the ram and a PSU for these reasons. But being an idiot, I left my old 6600GT at home that I generally use to verify if my card is shot..

Thanks for the input

would it make since that so far of my 4 sticks of ram, I have tried 3, one has passed almost 4 hours of memtest, and the other two, the monitor won't even turn on.

As for the fourth stick, I havent had time to try it since this problem began (its next) but I know from before the issue, that it has passed memtest.

could it be that two sticks are damaged?

Ya it could be ram

Could be RAM most definitely. I'm still, until you can eliminate it, thinking the video card is flaking on you. Ram usually will toss a BSOD or even beep at you. Whats making me think video is that when the system sets a while you seem to get a little video. After a few mins it seems you lose video but the machine is 'alive'. Here is a thought though. When a video GPU goes bad you'll usually get beeps from the board. If the memory ON the video card is dying it will produce artifacts like what you are seeing. At least that has been my experience.

Z

PS: Did the stick that passed always allow you to see video? As in, could you swap the one that passed in and every time it worked? OR did you run a test on the first on and the next 2 didn't get video? I ask because if you can place the one good stick in the system and all is well then I would obviously concur RAM is at least a part of this issue.

So I am not sure what was happening that the graphics card was responding. Also, I never really felt the need to keep the tiny internal speaker.. (I feel kind of dumb about that now).

Anywho, After switching completely back to my old board and still having the save problems, I let the system sit for a day in frustration and then came back, removed the cmos battery for 10 min and booted perfectly fine from one stick of ram. I immediately ran memtest and the stick passed. Still thinking it was my cpu or vga, I installed xp and ran OCCT and ATItool as described above. The VGA works but will artifact at times, so it will ahve to be replaced. The CPU has best every test so far. So, I concluded it must be ram and I somehow missed it with all this testing (maybe I didnt let cmos reset long enough between switch sticks.. idk) But I went back and tested each stick of ram individually and 2 prevented and response from the monitor and POST and two passed with flying colors.

So, something must have happened to those two sticks of ram, not sure what (although I suspect they were always faulty. As for the VGA, I think it has been artifacting for a while, I just haven't stressed it with testing or gaming to notice.

So anyways, I ahve moved back to the new, p45, mobo and everything is fine when running those 2 healthy sticks of ram.

Thanks for all the input guys. this was quite the odd and frustrating process.

I been having issues with my new rig that I built, my sig. And since I bought it recently I went throught a bunch of RMAs for replacements and got no where. On the 29th day of my return policy I decided to run memtest (should of done that first thing). And Found out that my 6gb (3x2gb) ram was at fault for all my issues. One stick gave 8 errors in different memtests, and another stick gave 750+ errors.

And all of my issues were my ram's fault even thought I returned my graphics card and got a different one since my first card was giving me nv4_disp.dll blue screens and I thought it was the card and not the ram. Im glad thought, I already thought that card was defective cause of the condition it came in and after returning it I was able to get the 5870 cause it just came out the next week after I initially bought the comp.

I was able to test my ram with different bios settings and nothing worked (since i read that altering bios settings could eliminate memtest errors). So I got my ram RMAed for replacement on the 30th day of the newegg policy. Since I was in a rush cause it was the last day, I didnt have the time to test each stick one at a time so it could be my mobo but with the various errors I got and the way it went about im almost 100% sure its the ram and had some backup from others that also agreed with it.


So what Im trying to say is, it could be the RAM since the errors ram can give you are very sketchy and you may end up thinking its other hardware. But with your memtesting and such, I believe that you should be fine.