Somehow I don't think mucking with drivers and environments will help.

Try stepping through the initializations that happen when the program is first launched, comparing the Z with either broken card. Somewhere some initialization is going wrong. A different approach might be to find all CUDA calls and put a printf after them for any errors whatsoever (presumably whatever is failing doesn't yet have an error check).

This is old version.(old normalize kernel)
Could you try ?

Dubslow is Best.
I will go sleep.:smile:

The old version gives residues on these cards.

Looking into further...

[QUOTE=Madpoo;425032]Are you the user who tried to manually submit two different "is prime!" results today?
LOL[/QUOTE]

It was pretty exciting to see

M( 45...... )P, offset = 32941289, n = 2592K, CUDALucas v2.05.1, AID: ...

but no, I did not submit it.

[QUOTE=Dubslow;425057]A different approach might be to find all CUDA calls and put a printf after them for any errors whatsoever (presumably whatever is failing doesn't yet have an error check).[/QUOTE]

This is good code to have present all the time.

[QUOTE=Dubslow;425057]. Somewhere some initialization is going wrong. [/QUOTE]

I don't know if this is a hint or not, mine is reporting a "clockRate" about twice the real clock.
Real clock = 575 reported clock = 1147

[QUOTE=bgbeuning;425070]It was pretty exciting to see

M( 45...... )P, offset = 32941289, n = 2592K, CUDALucas v2.05.1, AID: ...

but no, I did not submit it.[/QUOTE]

Well, there have been no less than 3 false positives (2 different users) in the past ~ 16 hours, all from CUDALucas 2.05.1 ...

Even though I'm totally sure they're false positives (2 are from double-checks, making it even more unlikely), I'm still running my own.

One of them is done and it's not prime of course... waiting on the others:
[URL="http://www.mersenne.org/M37182127"]M37182127[/URL]

[QUOTE=airsquirrels;425052] 690 is compute 3.0, 590 is 2.0,Z is 3.5.
[/QUOTE]

What compute capability did you compile CUDALucas for? (always check silly things first)

Chris

[QUOTE=msft;425049]I can read code.
What can i do ?:smile:[/QUOTE]

Please add code to exit with an error message if when you write a save file (or any other convenient time) you find that the LL iteration is zero or two.

[QUOTE=Prime95;425093]Please add code to exit with an error message if when you write a save file (or any other convenient time) you find that the LL iteration is zero or two.[/QUOTE]
I am not understand this issue.
Why iteration is only 0 or 2 ?
Please anyone explain it to me.