AMD 5850 BSDO & crashing had enough

[TrojanWhore]

is your CPU overclocked/volted at all? my friends [email protected] + 5850 heavily OCed used to put the hurt on his old 700w a little, and at 3 years old, I'd imagine your PSU has lost some of its nominal wattage.
some OCZ psus have a pretty bad rep, some are apparently ok, but I don't remember which models I'm afraid.
do you have access to another PC which you could test the card in, or possibly a highger rated PSU you could borrow off someone to determine where the problem lies?

 

[C64]

Try down clocking the card slightly say if it's at 725mhz by default try 715 mhz to me this sounds like a card that can no longer do it's clocks which can happen.

 

[Phixsator]

I was actually for a period having the exact same issues as the OP running on a Core I7 920 on a GHigabyte x58 UD4P motherboard. All sort of issues pointing towards memory. What it turned out to be was the motherboard.. I tested my OCZ ram for hours on end, tried other GFXs and the only things that was the same was the PSU, CPU, RAM and harddrives. Turned out the mainboard was bad. Changed that to a asus x58 sabertooth and my system has been singing ever since.

I know gigabyte is popular but im never ever buying that brand again(had 2 brand new gigabyte gtx 560 ti's fail on me at the same time).

So you might want if you have the chance to test another motherboard doesnt even need to be the same CPU as long as everything else is the same. Just a thought

 

[MeMeMeMe]

OP: I don't know why you think the 5850 has stability issues. it's one of the finest cards AMD/ATI have ever produced.

What it turned out to be was the motherboard..

I was just about to post this too:
You're having problems with RAM not being detected, with DVD drive not being detected, and with BSODs. That suggests motherboard to me.

What I'd suggest to rule out some other causes: If CPU or GPU are overclocked, reset them to stock. Then in bios, reset everything to optimum defaults.

Then run it for a few days and see if you still have problems.

 

[Dr Lodge]

Went to kick off memtest this morning before I went to work. It wouldn't get beyond 7, was exercising the 4gb-6gb region and looks like it just hung. Tried a few times and result was the same. Gonna try running it again just with 2x2gb modules and try and figure out whats going on.

Sounds to me like memory or motherboard, which would be a bugger. Above seems consistent with the earlier issue I had with 6gb being detected but only 4gb usable.

Guess if I can get 2x2gb memory working and all 3 memory modules work fine, then it must be motherboard right?

 

[Dr Lodge]

Results of memory test (I have 3x2gb sticks)
3 sticks in: freezes at 7% on first phase
2 sticks in, runs for a few seconds and at 47% through the first phase, reboots,
1 stick in - 2 sticks run fine, 1 stick has errors galore.
2 good sticks in - same as above reboots.

This is on default memory board settings (fail safe) except for memory voltages that are set to 1.35 and 1.65 volts.

So I think I have a duff memory board, and one duff memory stick, do you guys agree? I'm looking at a GA-EX58-UD3R board on Ebay which is quite cheap, so not much risk if it doesn't work or doesn't result in a better outcome. Assuming I win that, I can check what my PC is like with a new MoBo and 4GB RAM and take it from there.

 

[Dr Lodge]

OK this is really weird. Using the 2 good sticks in slots 1 & 3, the PC reboots after a short while when using memtest v3.5. But if using v3.4 it runs fine with no errors. I googled it and found some people have the same problem
http://forums.overclockersclub.com/index.php?showtopic=171697
http://forums.anandtech.com/archive/index.php/t-2042391.html
http://gskill.us/forum/archive/index.php?t-6879.html

Symptoms are - test a stick 1 by 1 on v3.5 and it works, but 2 or more sticks and it reboots around the 7% mark...try v3.4 and it works fine

 

[Dr Lodge]

Ran memtest v3.4 with all 3 memory sticks, the good sticks are in slots 0 and 1, bad stick in slot 2. Lots of errors showing in slots 0 and 1, so I assume this is related to the motherboard e.g. memory controller? Sorry for all the posts, just not sure what to make of all of this.

 

[MeMeMeMe]

Wht you hsould do is poick one stick of RAM, just one.
Test it in each slot. Record results.

Then pick the second stick, test it in each slot.

And continue till all are done.

It's more laborious but it's the only way to find out if:
one or more slots are faulty, or one or more ram sticks are faulty.

 

[Dr Lodge]

Wht you hsould do is poick one stick of RAM, just one.
Test it in each slot. Record results.

Then pick the second stick, test it in each slot.

And continue till all are done.

It's more laborious but it's the only way to find out if:
one or more slots are faulty, or one or more ram sticks are faulty.

I've kind of done that, although not exhaustively.

I've tested each of the 3 sticks in slot 1 - sticks 1 and 2 passed, stick 3 failed.

I then used one of the good sticks and tested in slots 1, 3 and 5 (the ones I use). All passed.

I'm now using sticks 1 and 2 in slots 1 and 3. I got a couple of errors last night on the 2nd pass, but then I re-ran overnight and it did 6 passes clean, that was on fail safe defaults. I've now used optimised defaults and am re-running. This is all using memtest v3.4 as v3.5 won't play ball.

 

[Dr Lodge]

Figured that since my existing memory is not getting through memtest without errors, I would replace it and if the motherboard is faulty, then new memory isn't wasted either.

Got 4x4gb sticks from crucial (ballistix tactical), it all gets through memtest OK and so far no blue screens. Fingers crossed.

 

[Dr Lodge]

Added another 8gb of memory, so banks are full with 24GB and using triple channel. Memtest passes all OK, and no blue screens either. Looks like it was bad memory in the end, nothing to do with the graphics card.

 

[foil]

There would also have been a BSOD report in event viewer telling you what went wrong.

 

[Dr Lodge]

I looked for the BSOD reports, but I really couldnt find anything meaningful. I configured the PC to halt on BSOD, and the error message there didn't tell me much either.