[QUOTE=owftheevil;427967]Correction made. It now aborts the test when a premature 0 residue occurs. It wil take a few days to get sourceforge updated. (Extemely unreliable and slow internet connection.)[/QUOTE]

Thank you owftheevil, let us know when the new build is ready for download :-)

[QUOTE=owftheevil;427967]Correction made. It now aborts the test when a premature 0 residue occurs. It wil take a few days to get sourceforge updated. (Extemely unreliable and slow internet connection.)[/QUOTE]

Thanks. :smile:

When a new prime is reported manually, we get emailed about it and it takes a little bit to say "wait a minute, this was only assigned a couple hours previously"... LOL

[QUOTE=Madpoo;428104]Thanks. :smile:

When a new prime is reported manually, we get emailed about it and it takes a little bit to say "wait a minute, this was only assigned a couple hours previously"... LOL[/QUOTE]

So you're saying if I want to plan an April fool's joke, I should reserve the exponent today?

[QUOTE=Madpoo;425083]Well, there have been no less than 3 false positives (2 different users) in the past ~ 16 hours, all from CUDALucas 2.05.1 ...[/QUOTE]

Sigh... another one today. Just got the assignment earlier in the day and keeps trying to manually submit their "Mxxx is prime!" result. It (probably) isn't, CudaLucas 2.05.1 apparently has a bug that reports it as prime within hours (minutes?) of starting.

Walk Through Install
 

Hello,

I'm attempting to install CUDALucas on my Ubuntu 14.04 system. Read the README, yet still I am very confused on how to actually install the program and get it up and running. I have the pre-recs installed (gcc & CUDAToolkit). Any help would be greatly appreciated.

I'd like to improve my GTX 1xxx data for [url]http://www.mersenne.ca/cudalucas.php[/url] but I'm lacking benchmarks.
Can anyone who has a GTX 1xxx card please send me a benchmark as described at the top of the above link.
I've seen evidence that performance can be severely limited if not using PCIe 3.0 x16 slot, so please also confirm the speed and generation of the slot your GPU is running in when submitting benchmarks.

Here's some:
[CODE]| Aug 04 16:19:43 | M57885161 900000 0x0000000000000000 | 3920K 0.00000 0.0468 4.68s | 43:01 1.55%
| Aug 04 16:19:48 | M57885161 1000000 0x0000000000000000 | 3920K 0.00000 0.0464 4.64s | 43:03 1.72%
| Aug 04 16:19:52 | M57885161 1100000 0x0000000000000000 | 3920K 0.00000 0.0476 4.76s | 43:10 1.90%
| Aug 04 16:19:57 | M57885161 1200000 0x0000000000000000 | 3920K 0.00000 0.0446 4.46s | 43:01 2.07%
| Aug 04 16:20:01 | M57885161 1300000 0x0000000000000000 | 3920K 0.00000 0.0442 4.42s | 42:51 2.24%[/CODE]
This is 1080 at 1.683GHz and PCI-E v.3.
Rubbish fp64 performance and they say Titan X (P)'s fp64 is relatively the same.

[QUOTE=Karl M Johnson;439315]This is 1080 at 1.683GHz and PCI-E v.3.[/QUOTE]That seems slower than expected. Is it really running at 3920K FFT instead of the usual 3136K?

[QUOTE=James Heinrich;439317]That seems slower than expected. Is it really running at 3920K FFT instead of the usual 3136K?[/QUOTE]

Might be the non-ideal FFT size. My 1080 is currently working M76042667 at a consistent 3.65ms/iteration (4096K FFT).

[QUOTE=airsquirrels;439342]My 1080 is currently working M76042667 at a consistent 3.65ms/iteration (4096K FFT).[/QUOTE]Clock speed, please, so I can add that as another data point.

[QUOTE=James Heinrich;439347]Clock speed, please, so I can add that as another data point.[/QUOTE]

Shows 1715 MHz currently, stock FE