• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Twinmos 3200 512meg Memtest 3.0 errors

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Aviator747

Member
Joined
Jun 7, 2002
Location
Cornwall, Ontario, Canada
All you guys that are running 200+ fsb on a NF7-S motherboard have you tried running memtest to see if you have any errors?

I bought a stick of Twinmos 3200 512meg BH5 from a guy on amdmb who was dealing them, back at the end of march. It took me a couple of weeks to accquire a NF7-S 2.0. I tinkered & puttered around with my new toys, oc'd my chip.

I ran mem test 3.0 & would always run into errors on test #5. The errors stopped & then returned. I tinkered & puttered some more, flashed the bios several times. I upgraded my PSU to a Antec true power 480, as my generic was causing me some grief when oc'ing.

My box seems pretty stable. But when I run memtest 3.0 I get a boat load of errors from the lower meg range all the way up to 512. It always seem to be on test #5. It has also happened on one of the higher number tests too. I am not talking about a few errors either, I am talking a few hundred thousand errors.

This is at any cas, ras rating, cpu interface enabled\disabled. I had to hit it with 2.9v to get it stable at 200 mhz in windows & to also get it fairly stable in prime95 "ver 23.5.1". This is at 6/6 synced fsb/ram. I am running bios 18. I am also running me VDD at 1.7. Also I don't have any form of cooling on my south bridge.

I also get this error which reads something like "Could not write to alocated memory path" or something like that. I should have written that message down. That is also another reason why I bumped my vdimm up.

So anyone have any suggestion, ideas? Should I seriously start looking at RMA'ing?

Thanks for any info guys. :)

I'm sorry for the lengthy post guys :eek:
 
Hmm, I'm not familair with using memtest. I've not used it since
it was beta, but that does sound like a bad stick of memory.

Do you have one stick or two? If you have two, test them
seperately and see if one stands out as bad. If you have only
one, test the system overall with other programs like
Prime95 (which usually shows errors rather quickly). If the
system is stable running 3d benchmarks and prime95 (for
several hours on end), then I'd have to question Memtest.

Oh, and long posts are good, it's a sign that you thought out
your question and laid it out well so people and understand
it and answer it better. I wish more people wrote like you
and I!!
 
I am using one stick of ram. I had run prime95 several times before & it would always fail on me. At first after the first several tests. Then I got it to run longer & hadn't played with it anymore.

So last night afer I posted my message, I fired up prime95 & it is still running fine as of this posting. I have also run 3dmark2001se succesfuly previously.

I also posted my problems in amdmb & the person who sold me the ram & a few other people responded. They said with the newer nforce boards have a problem with test #5 & it is quite common to see errors on this test. They also asked how warm my case was. They said high case temperatures seem to make this problem worse.

EDIT: Here is a picture of that error message that I get at times as stated above. http://www3.sympatico.ca/lauzon_c/error.JPG
 
Last edited:
Back