Massive memory test failure on otherwise stable system. Confused.

[poco]

I just got 4 sticks of Kingmax PC4000 DDR and was eager to try them out,
and it's not exactly turning out to be a bundle of fun I hoped.
First, I ran memtest+ 1.27, and it kept freezing in certain tests depending
on what clock speeds I ran them at. I even underclocked them to 190, but it
still froze up. changed vcore, ddr voltage, same result. Tested each
sticks, tried them in different slots, and memtest+ still freezes. Tried
memtest+ 1.20, which was the version I last successfully ran memtest+ with
Geil PC3200 I had previously, and still same thing.
I then tried Windows Memory Test, now it gives errors in every 4th
address. I tried them in different settings and individually, but still
same errors.
I never have seen anything like that, so I decided to see how XP
actually behaves. I booted to XP and ran Sandra while OCed to 225Mhz(P4
clocked to 3.6 from 3.2 with FSB of 900 at 1:1 setting). The system seems
to be stable. Sandra benches right where it should, and I've been running
Prime95 for past hour or so with no errors.
I have to go to class in couple hours, so I'm going to take a quick
nap. I'll leave Prime95 running to see what happens, but I'm really
confused now and was hoping someone can help me out. Thanks for all the
helps.

Oh here's my system.
ASUS p4c800E deluxe
P4C 3.2 @3.6
Kingmax PC4000 @ 1:1
9800XT stock speed
Thermaltake 480W

 

[JAD]

you built your machine to test memory? prime example of bogus
'software' memory testers....

 

[Matt]

First, I ran memtest+ 1.27, and it kept freezing in certain tests

I've been having a similar problem with a Dell Dimension 4600 w/ 2.4G P4
and Intel 845 chipset. Memtest hangs, but Windows and Linux seem to run
fine.

 

[Shep©]

I've been having a similar problem with a Dell Dimension 4600 w/ 2.4G P4
and Intel 845 chipset. Memtest hangs, but Windows and Linux seem to run
fine.

As I've stated many times.Software RAM test programs are not worth the
download let alone the running.How does the user know the program has
been well written?
How does the user know that even if it is well written it has taken
into account the different maker's tolerances and RAM configurations
on-chip?
How does the user know that the download was not corrupted?
There are many variables that add to their built in faults.

Windows(cannot speak for Linux) is in itself a good memory tester as
it will fall over giving,"Registry Errors" and corrupt File downloads
if the RAM is not up-to-par.



--
Free Windows/PC help,
http://www.geocities.com/sheppola/trouble.html
remove obvious to reply
email (e-mail address removed)
Free original songs to download and,"BURN" :O)
http://www.soundclick.com/bands/8/nomessiahsmusic.htm

 

[David Maynard]

As I've stated many times.Software RAM test programs are not worth the
download let alone the running.How does the user know the program has
been well written?
How does the user know that even if it is well written it has taken
into account the different maker's tolerances and RAM configurations
on-chip?

And just how do you know that 'windows' (sic) "is well written it has taken
into account the different maker's tolerances and RAM configurations on-chip??

How does the user know that the download was not corrupted?

The same way you know whether *any* downloaded program was 'corrupted' ornot.

There are many variables that add to their built in faults.

Just asking "how do you know?" doesn't constitute 'built in faults'.

Windows(cannot speak for Linux) is in itself a good memory tester as
it will fall over giving,"Registry Errors" and corrupt File downloads
if the RAM is not up-to-par.

Lots of things cause windows to fall over. That doesn't tell you what did
it, nor does windows running ok 'right now' mean the hardware is fine.