[QUOTE=Madpoo;426043]Meanwhile, if you're hungry for some good old fashioned strategic double checks, here's the latest batch of 2+ to 1 bad-to-good. Some are in the cat 4 "churn" category (the 40M-42M) so get 'em while you can.

[CODE]exponent Bad Good Unk Sus Solo Mis worktodo
38341909 2 1 2 0 2 0 DoubleCheck=38341909,71,1
38456843 2 1 2 0 2 0 DoubleCheck=38456843,71,1
38898667 8 4 3 1 2 2 DoubleCheck=38898667,71,1
40459187 5 1 2 0 2 0 DoubleCheck=40459187,72,1
40494269 6 0 1 0 1 0 DoubleCheck=40494269,72,1
40575707 2 0 3 0 1 2 DoubleCheck=40575707,72,1
41537261 4 0 1 1 1 1 DoubleCheck=41537261,72,1
42393961 2 0 5 0 5 0 DoubleCheck=42393961,72,1
42430061 2 1 2 1 2 1 DoubleCheck=42430061,72,1
52276529 21 3 5 5 5 5 DoubleCheck=52276529,73,1
58080709 2 0 4 0 4 0 DoubleCheck=58080709,73,1[/CODE][/QUOTE]

I took all these except the ones already grabbed by the churners.

[QUOTE=Madpoo;425404]
[CODE]exponent Bad Good Unk Sus Solo Mis worktodo
51753259 3 0 2 1 2 1 DoubleCheck=51753259,73,1
[/CODE][/QUOTE]

I've taken this one.

Quote:
[SIZE=2]Originally Posted by [B]kladner[/B] [URL="http://www.mersenneforum.org/showthread.php?p=425858#post425858"][IMG]http://www.mersenneforum.org/images/buttons/viewpost.gif[/IMG][/URL] [/SIZE]
[SIZE=2] [/SIZE][I][SIZE=2]Thanks. I may have to keep that 580 doing TF. It does not seem to be totally reliable on LL.[/SIZE][/I]

[QUOTE=0PolarBearsHere;425934]Yeah I gave up on LLing with my 770 for the same reason.[/QUOTE]

On this theme, I am trying an approach which I have talked about, but never followed through on. I am running a regular DC on the 580, without using a looping batch file to start it. I have the VRAM slowed from 2100 to 1900 MHz. This is in line with the way I have set up to run CuLu on other cards.

If CuLu times out, it will just stop. I may loose some processing time this way, but I will be able to reset the memory clock to the lower speed after a driver reset.

I started this run early this morning. I am delighted that it has now continued for a little over six hours without stalling. Of course, verified runs don't prove reliability, but they would increase my confidence in the settings for the card.

[QUOTE=kladner;426223]Quote:
[SIZE=2]Originally Posted by [B]kladner[/B] [URL="http://www.mersenneforum.org/showthread.php?p=425858#post425858"][IMG]http://www.mersenneforum.org/images/buttons/viewpost.gif[/IMG][/URL] [/SIZE]
[I][SIZE=2]Thanks. I may have to keep that 580 doing TF. It does not seem to be totally reliable on LL.[/SIZE][/I]



On this theme, I am trying an approach which I have talked about, but never followed through on. I am running a regular DC on the 580, without using a looping batch file to start it. I have the VRAM slowed from 2100 to 1900 MHz. This is in line with the way I have set up to run CuLu on other cards.

If CuLu times out, it will just stop. I may loose some processing time this way, but I will be able to reset the memory clock to the lower speed after a driver reset.

I started this run early this morning. I am delighted that [U]it has now continued for a little over six hours without stalling.[/U] Of course, verified runs don't prove reliability, but they would increase my confidence in the settings for the card.[/QUOTE]

Now 16.5 hours uninterupted, 58.09% complete, ETA 12 hours neat. I hope these updates don't jinx the process, or unduly provoke the Noodly Appendages. :sirrobin:

[QUOTE=kladner;426271]Now 16.5 hours uninterupted, 58.09% complete, ETA 12 hours neat. I hope these updates don't jinx the process, or unduly provoke the Noodly Appendages. :sirrobin:[/QUOTE]

I might look into that. My problem is that I'll go back and see that CudaLucas has quit and I need to restart it. I just took to making a bat file to run it about 30 times in a row to help counteract it.

[code]DoubleCheck=46189607,72,1[/code]

No match, [COLOR="Red"]unverified[/COLOR].

[QUOTE=kladner;426271]Now 16.5 hours uninterupted, 58.09% complete, ETA 12 hours neat. I hope these updates don't jinx the process, or unduly provoke the Noodly Appendages. [/QUOTE]
The run finished with one timeout. Result matched. This was a regular DC, not a strategic check against machines with bad histories. It is more useful at the moment to test the setup where there is a better chance of a match if the result is correct.

Mismatch for 51770489, could someone do the TC?

Mismatch for 51940249 - I suspect user 'lelikoid' had a rather flaky computer at some point.

[QUOTE=fivemack;426540]Mismatch for 51940249 - I suspect user 'lelikoid' had a rather flaky computer at some point.[/QUOTE]
I will take both from your last 2 posts.

Anyone mind running a quad-check on 36122771? I've been using triple checks to verify stability on a bunch of GPUs and that one came mismatched.