P4C800-E Deluxe Memory Problem

[Donald C. Stuart]

I have a P4C800-E D mobo, with two Corsair CMX512-3200 C2 DDR dimms
set at 2-3-3-6. The system lately has taken to crashing several times
a day, and I assumed some driver conflict. Recently, though, Windows
has been sending back event messages claiming a memory problem, and
offering a free memory diagnostic download. Sure enough, if I run the
Windows diagnostic, the "LRAND" test fails (although all other tests
pass) This is a random number write read, and the screen suggests that
memory read is failing several times per test--but always the same
fail. The test reads 1 (integer) less than what was written. When I
take out one memory stick, every seems to work and test good whichever
stick I put back in. What might cause a dual-stick error read? Is this
an ASUS or a Corsair problem? Many many thanks for any ideas!

 

[Ghostrider]

I have a P4C800-E D mobo, with two Corsair CMX512-3200 C2 DDR dimms
set at 2-3-3-6. The system lately has taken to crashing several times
a day, and I assumed some driver conflict. Recently, though, Windows
has been sending back event messages claiming a memory problem, and
offering a free memory diagnostic download. Sure enough, if I run the
Windows diagnostic, the "LRAND" test fails (although all other tests
pass) This is a random number write read, and the screen suggests that
memory read is failing several times per test--but always the same
fail. The test reads 1 (integer) less than what was written. When I
take out one memory stick, every seems to work and test good whichever
stick I put back in. What might cause a dual-stick error read? Is this
an ASUS or a Corsair problem? Many many thanks for any ideas!

We use Kingston Technologies RAM and its recommended timings
are 3-3-3. We have had no problems at this timing nor 2.5-3-3
nor via SPD. The ASUS manual configured settings are 2.5-4-4-8.
The Corsair settings here seem a bit aggressive.

 

[Paul]

I have a P4C800-E D mobo, with two Corsair CMX512-3200 C2 DDR dimms
set at 2-3-3-6. The system lately has taken to crashing several times
a day, and I assumed some driver conflict. Recently, though, Windows
has been sending back event messages claiming a memory problem, and
offering a free memory diagnostic download. Sure enough, if I run the
Windows diagnostic, the "LRAND" test fails (although all other tests
pass) This is a random number write read, and the screen suggests that
memory read is failing several times per test--but always the same
fail. The test reads 1 (integer) less than what was written. When I
take out one memory stick, every seems to work and test good whichever
stick I put back in. What might cause a dual-stick error read? Is this
an ASUS or a Corsair problem? Many many thanks for any ideas!

Get yourself a copy of memtest86+ from www.memtest.org .

The difference is: memtest86+ is a standalone test program that
boots from its own floppy image. No OS is present while it tests,
and in fact the program moves itself out of the way, and
"tests underneath". So, every pass done, tests the whole memory.

A Windows test program, can only test the memory not occupied by
the OS.

You have several test cases to run. You can run in dual channel
mode and verify your memory. Note the address of the failure
(but you'll have fun figuring out which stick that address
actually lives in). Next, test one stick at a time. You could
test one stick at a time on Channel 0 and then one stick at
a time on Channel 1, if you think this is a Northbridge or
particular memory channel issue.

Another thing I recommend, is getting a copy of CPUZ from
cpuid.com , and verify the frequencies and memory timimg
parameters being used by the system. The Asus BIOS doesn't
always set the system up correctly, so a quick check with
CPUZ wouldn't hurt.

The first (crap) memory I used on my system, failed memory
test at the same location in memory each time. That made
it pretty easy to decide the memory was bad (and out of
warranty). Using loose timing parameter with my memory,
didn't help a bit, so it was a more-or-less stuck fault.

Once you resolve this issue, one way or another, a run
with Prime95 (torture test option in the menu) from
mersenne.org, will tell you whether memtest86 has found
all the problems or not. Prime95, or SuperPI, 3DMark, or other
programs like that, should be run to prove the computer
is ready for serious work. Waiting for Windows to crash
is a slow way to prove a system is stable

Paul

 

[Marcel Overweel]

I have a P4C800-E D mobo, with two Corsair CMX512-3200 C2 DDR dimms
set at 2-3-3-6. The system lately has taken to crashing several times
a day, and I assumed some driver conflict. Recently, though, Windows
has been sending back event messages claiming a memory problem, and
offering a free memory diagnostic download. Sure enough, if I run the
Windows diagnostic, the "LRAND" test fails (although all other tests
pass) This is a random number write read, and the screen suggests that
memory read is failing several times per test--but always the same
fail. The test reads 1 (integer) less than what was written. When I
take out one memory stick, every seems to work and test good whichever
stick I put back in. What might cause a dual-stick error read? Is this
an ASUS or a Corsair problem? Many many thanks for any ideas!

I had the same problem as you described.
Same board, different memory.

The first few weeks I didn't do very much with my system, just installing
programs and little filecopying stuff andso. No problems at all.
Later started running some games.
Once in a while a crash. Irritating but not enough go give me the feeling
that there was a hardware problem. I excpected it was because of new
graphics driver.
Then I started to extract some large rar archives: random crc errors, but
only after the computer has been running for 15 or so minutes.

Was advised to run memtest86, and also as you described, all tests were ok
except one.
Testing with one module always gave good results, didn't matter which module
I used.

Switched from two Cas2 to two Cas3 modules, problem solved.

regards,
Marcel