Help me troubleshoot a new system

[Tsumari]

I recently bought some parts for an Athlon XP 2500 based system. I'm using an NF7-S and Corsair XMS PC2700 at 2-2-2-5-T1 timings. Everything is running at stock speed.

I've had a couple of problems. First, when I initially put the system together it would hard crash pretty frequently, enough to corrupt the OS a few times and force a reinstall. Well I read up on the board and found out that I needed to connect a second power cord to the board, which I did. That appeared to solve my instability, but since then I've had errors in both Prime95 and Memtest.

Here is a snippet from my Prime95 error log:

[Fri Jan 02 14:13:32 2004]
Iteration: 1999090/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 4.701452292438471e+016 != 4.699693073834029e+016
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:18:40 2004]
Iteration: 2010387/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 6237645932838195 != 6272830304927026
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:25:41 2004]
Iteration: 2011680/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 840343478093143 != 769974733915488.4
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:31:01 2004]
Iteration: 2023461/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 1.044580371339811e+016 != 1.041061934130926e+016
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:36:35 2004]
Iteration: 2023805/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 6.709325823944098e+016 != 6.71284426115298e+016
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:43:44 2004]
Iteration: 2038819/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 4.557647276729898e+016 != 4.559406495334341e+016
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 14:54:16 2004]
Iteration: 2054319/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 74651240251616.72 != 57059054207180.85
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

[Fri Jan 02 15:00:09 2004]
Iteration: 2061426/10309567, ERROR: SUM(INPUTS) != SUM(OUTPUTS), 1.445014599752851e+017 != 1.444310912311074e+017
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.

My CPU is running at a temp of 44C (stock speed) and system temp is 27 C, so that seems pretty reasonable.

Memtest reports errors in my memory. I dont have a floppy drive installed so I haven't tried Memtest86 yet.

Also, occasionally C&C Generals will crash. Now, I am modding that game so I put about a 30% chance that my mod crashed it, but it has run perfectly stable on my 1ghz athlon in the other box, so I dont want to blame my modding just yet.

First, I am going to take the system apart and put it back together, then look to see if I still get the errors.

If I still get the same errors, should I just RMA my memory since Memtest reported errors, or should I RMA the CPU as well since Prime95 reported various errors?

The only expansion card I have installed is my radeon 7000 AGP, so I doubt a card is doing it. Aside from that, I have an old 24x cdrom and a WD "Special Edition" drive installed.

Let me know what you think.

[EDIT] changed incorrect memory timing info in post.

Tsumari

 

[TheMostWantedPolishTwin]

my call - memory... if it's possibe test it in ohter mobo...

 

[Tsumari]

My house has a lot of PCs, but this is the only system that has been built fairly recently. My last system is a 1ghz athlon, my brother has a 933 athlon, my mom a 600 athlon, etc.

I'm going to keep looking for benchmarks and stability tests to see if I can rule out the CPU. I need a way to try to verify the motherboard's stability but without a modern secondary PC, that will be hard.

Tsumari

 

[Tsumari]

I've got a new error. On reboot: "One of the files containing the system's registry data had to be recovered by use of a log or alternate copy. The recovery was successful."

Time for me to get cracking on some testing. I'll update in a while when I've pulled it apart and put it back together.

Tsumari