Can ANYONE with a P4C800-E Dlx Run Prime95 v23.7 Torture Test longer than 4 hours? FIXED! Prime code was at fault

[Amused]

Can anyone with an Asus P4C800-E Deluxe run Prime95 Blended torture test longer than 3 hours?

In the P4C800/E thread we seem to have found a bug that causes Windows to shut down P95 at about three or so hours when it reaches the M420217 with a 20K FFT test.

We have three people who are experiencing this problem.

Anyone else?

Just to note, we all have rev2.0 and BIOS 1011 or 1014.

Intel Chipset drivers v5.0.2.1003. Intel Pro1000 NIC drivers v7.0.37.0 are also common among us. Just on a whim, I will be trying the newer LAN drivers to see if that fixes it.

It seems to me few if any people seem to run prime very long if this bug has never been found before.

Please let us know. Just run prime while you sleep if you don't have the time. The bug simply causes Windows to popup a message saying there has been a problem with Prime95 and it must shut down. It does not cause a catastrophic failure or anything.

EDIT: This is ONLY with the newest version of Prime95. Version 23.7

Previous versions do not seem to have this problem.

 

[Amused]

FIXED!!!

The creater of Prime95 has siad this is an issue with the program, not our hardware and has issued a patch! See this thread at the Prime Forums:

http://www.mersenneforum.org/showthread.php?s=&postid=15801#post15801

 

[Amused]

I'm really surprised there isn't more interest in this.

 

[Thor86]

Originally posted by: Amused
I'm really surprised there isn't more interest in this.

What's the benefit of using the latest version? I just use previous version, no problems.

 

[Amused]

Originally posted by: Thor86

Originally posted by: Amused
I'm really surprised there isn't more interest in this.

What's the benefit of using the latest version? I just use previous version, no problems.

None, really. Although 23.7 and 23.8 run much faster than 23.6. But that only matters if you're using it to benchmark or run Prime95 DC. Although, it is better to run through the tests faster when stability testing because you get a broader range of tests in a shorter amount of time.

Anywho, 23.6 and previous versions are fine and dandy for stability testing.

My whole point was that 23.7 can cause Windows errors on some hardware... and that this is a software bug, NOT the fault of the hardware.

Very useful info for folks getting errors and tearing their hair out trying to figure out what's wrong with their rig.

 

[Amused]

Bump

 

[Jack4KickAss]

Haven't tried to run Prime 95 for that long, yet. Thanks for picking this up.

My PC3200 memory seem to not like any overclocking. Can't set my board to turbo, or run slightly tighter ram timings without a system becoming unstable.

Main reason for me not running Prime that long yet.

 

[Amused]

Originally posted by: Jack4KickAss
Haven't tried to run Prime 95 for that long, yet. Thanks for picking this up.

My PC3200 memory seem to not like any overclocking. Can't set my board to turbo, or run slightly tighter ram timings without a system becoming unstable.

Main reason for me not running Prime that long yet.

Have you tried upping the DDR voltage? My Mushkin PC3500 wouldn't run at 3200 speeds with tight timings until I upped the voltage to 2.75.

 

[Amused]

bump

 

[KF]

 

[KF]

 

[KF]

 

[KF]

>A geek doesn't have to be a fat slob or a skinny wimp.

True. You can also be skinny slob or a fat wimp.

Why would you want to waste time working out when you could be putting that time to good use experimenting with computer hardware?

 

[Amused]

Originally posted by: KF
>A geek doesn't have to be a fat slob or a skinny wimp.

True. You can also be skinny slob or a fat wimp.

Why would you want to waste time working out when you could be putting that time to good use experimenting with computer hardware?

Because I haven't found a way to make my computer extend my life, or provide the needed components for procreation.