• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

RMA an AMD Athlon64 3200+? (UPDATE!!)

D

deveraux

Senior member
I recently bought all the components for my new rig and set it up. Encountered some problems with the RAM modules but solved it by increasing the Vdimm to 2.8V and is memtest stable for ~6-8 hrs (I manually stopped the test since I needed to use the system).

However, this system seems to fail Prime95's in-place large FFT after 2-3 hrs (3:38 longest, 2:17 shortest), but can run small FFT for > 10 hrs. I don't think its the RAM since I not only ran all the tests but also ran tests 5 & 6 individually for ~ 3 - 4 hrs each while getting my RAM stable at the Vdimm of 2.8V.

Although its not Prime95 stable, I don't notice any system instability during normal operation except for the occasional IRQ_LESS_THAN BSOD when I exit UT2004 but I'm pretty sure that that's software related.

So, the big question is should I call for an RMA of the CPU since it doesn't appear to be unstable during "normal" operation? What do you guys think?

Thanks!
 
Specs:

AMD Athlon64 3200+
2x512 MB Corsair VS RAM
DFI NF4 UT LanParty SLI-D
Leadtek GeForce 6800GT 256MB
Creative SoundBlaster Audigy2 ZS

WD Raptor 74GB
Maxtor DiamondMax 10 300GB

ASUS DVD-ROM
NEC 3520A DVD-RW

Antec NeoPower 480W

Don't think I missed anything else.
 
let me know if i'm wrong, but usually if there's something wrong with the cpu the system won't function. are you sure it isn't still a ram issue?
 
RAM timings are set to SPD. According to the POST screen, its: 2.5 - 3 - 3 - 8, @ 1T. I seem to remember about a month or two ago that some people with AMD64's were failing prime at stock as well, any chance this is related?

fishmonger: As far as I'm aware, memtest thoroughly stresses the RAM, hence why I don't think its the RAM, but for all I know, you could be right. Any way of testing out that theory?
 
"2x512 MB Corsair VS RAM "

trust me, i've used value ram on that mobo you have and it gave me the same problems

i switched ram and everything was perfect again

don't get value ram
 
So its the RAM then? Well, that's annoying, since it passed memtest, I don't think I can even RMA it. Oh well, maybe I'll leave memtest on for 24hrs and hope that it errors so I have a good excuse to RMA it. Or else, I'll just have to cough up the dough I suppose.
 
let me add this:

I got a friend a 512 stick of kingston for their pc to add to their 256 meg stick of nanya ram.

it went through 13 hours of memtest(only stopped because they wanted to use their pc)

but they complained over problems and whatnot. turned out it couldn't pass more than 5 minutes of prime

i took out the 256 stick at the very end and BAM....it went through 4 hours prime

personally i memtest for 72 hours and prime for a week. if i can do that, PC is good to go
 
Have you tried running Prime 95 with timings at 2.5-4-4-8-2T? If that's stable, then I would think it's safe to eliminate the CPU as the culprit.
 
Originally posted by: diabloII
"2x512 MB Corsair VS RAM "

trust me, i've used value ram on that mobo you have and it gave me the same problems

i switched ram and everything was perfect again

don't get value ram
Click to expand...

there's nothing wrong with value ram, you just need to make sure that it isn't the problem. Memtest would definately be a way to see if your ram is the problem.
 
loafbred: Ok, I will run Prime95 with those memory timings tonight and post back here in the morning.

Until then, anyone else has any other opinions/ideas about this?
 
Dont blame the hardware if your software installations are not installed correctly or if you have unstable junk programs installed....
 
If you don't have any problems except when you run Prime95, then I don;t see the problem. Just don't run Prime anymore.
 
Originally posted by: Ike0069
If you don't have any problems except when you run Prime95, then I don;t see the problem. Just don't run Prime anymore.
Click to expand...

Very BAD advice. If prime's got a problem, you got a problem.

'Oh, my engine is making funny noise. I'll just ignore it and pretend like nothing's wrong'
 
Set your memory timings to 3-4-4-8 and run the Prime95 Large FFT torture test. If it fails bump the cup core voltage from 1.40 to 1.425 and run it again.
 
Since your RAM is fine, I'm thinking you might have one of those earlier Winchesters that had problems with P95.
 
I agree for the most part with everything said in this thread. Could be CPU or RAM. I am concerned that at SPD the RAM is at 1T; I wouldn't think that value select would default to 1T. I'm am mostly interested in this topic because I just purchased the exact same RAM. Hope I don't have the same problems.
 
Originally posted by: Guild
I agree for the most part with everything said in this thread. Could be CPU or RAM. I am concerned that at SPD the RAM is at 1T; I wouldn't think that value select would default to 1T. I'm am mostly interested in this topic because I just purchased the exact same RAM. Hope I don't have the same problems.
Click to expand...

I have rechecked the BIOS settings, and Command Per Clock (CPC) is indeed set to Auto and hence SPD (if I'm not wrong) and it still sets the RAM sticks to 1T.

I have now set my RAM timings to 3-4-4-8 and am running Prime95. Will post back if problem still persists.

Originally posted by: John
Set your memory timings to 3-4-4-8 and run the Prime95 Large FFT torture test. If it fails bump the cup core voltage from 1.40 to 1.425 and run it again.
Click to expand...

Doesn't bumping core voltage void warranty? My current Vcore is 1.38 according to BIOS and 1.36 according to CPU-Z.

Originally posted by: Garlic
Dont blame the hardware if your software installations are not installed correctly or if you have unstable junk programs installed....
Click to expand...

I wasn't blaming UT2004 for crashing as I don't believe the hardware has anything to do with it. But I am blaming the hardware for not being able to run Prime95 for at least 12 hrs without giving an error.
 
If you have any intention of ever being able to RMA your CPU honestly, do not overclock it or raise vcore.

That being said, it looks like it's either an issue with your RAM, or potentially the on-chip memory controller.

Also, have you tried testing with only one stick in single-channel mode?
 
Originally posted by: DrMrLordX
If you have any intention of ever being able to RMA your CPU honestly, do not overclock it or raise vcore.
Click to expand...

WTF are you talking about?!
Obviously overclocking voids warranty, and once you oc'ed you shouldn't RMA your chip. That said, there's no way the vendor/manufactorer can tell that the chip was overclocked, unless you were a totall moron.
Further more, not all systems are EXACTLY the same, and some systems, might give you 1.4v in the Vcore, while in other systems, this 1.4v might actually be slightly higher, due to different mobo/psu/dust/whatever, that deviation from standard doesn't necesarily means you overclocked. So, unless you were pumping 1.7v and fryed your chip when you're supposed to use 1.4v, there's really no telling if you overclocked or not.

:cookie:
 
Originally posted by: DrCrap
Originally posted by: DrMrLordX
If you have any intention of ever being able to RMA your CPU honestly, do not overclock it or raise vcore.
Click to expand...

WTF are you talking about?!
Obviously overclocking voids warranty, and once you oc'ed you shouldn't RMA your chip. That said, there's no way the vendor/manufactorer can tell that the chip was overclocked, unless you were a totall moron.
Further more, not all systems are EXACTLY the same, and some systems, might give you 1.4v in the Vcore, while in other systems, this 1.4v might actually be slightly higher, due to different mobo/psu/dust/whatever, that deviation from standard doesn't necesarily means you overclocked. So, unless you were pumping 1.7v and fryed your chip when you're supposed to use 1.4v, there's really no telling if you overclocked or not.

:cookie:
Click to expand...

Oh sh!t, I just realized you said HONESTLY....
then you're right... 😱
 
Originally posted by: DrCrap
Originally posted by: DrCrap
Originally posted by: DrMrLordX
If you have any intention of ever being able to RMA your CPU honestly, do not overclock it or raise vcore.
Click to expand...

WTF are you talking about?!
Obviously overclocking voids warranty, and once you oc'ed you shouldn't RMA your chip. That said, there's no way the vendor/manufactorer can tell that the chip was overclocked, unless you were a totall moron.
Further more, not all systems are EXACTLY the same, and some systems, might give you 1.4v in the Vcore, while in other systems, this 1.4v might actually be slightly higher, due to different mobo/psu/dust/whatever, that deviation from standard doesn't necesarily means you overclocked. So, unless you were pumping 1.7v and fryed your chip when you're supposed to use 1.4v, there's really no telling if you overclocked or not.

:cookie:
Click to expand...

Oh sh!t, I just realized you said HONESTLY....
then you're right... 😱
Click to expand...

EDIT button 😀
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top