mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   GPU Computing (https://www.mersenneforum.org/forumdisplay.php?f=92)
-   -   LLRCUDA - getting it to work (https://www.mersenneforum.org/showthread.php?t=18644)

diep 2013-10-02 00:54

LLRCUDA - getting it to work
 
hi!

Trying to get the Tesla's here going with LLR CUDA.

After some attempts got to work LLR CUDA at a box.
Didn't figure out yet how fast it is.
Seems to work for 3 * 2 ^ n - 1 here.

I'd like to test 69 * 2 ^ n - 1 with it.

Maybe i have the wrong syntax?

./llrCUDA -q"69*2^2410035" -d

Then it reports "Not support zero-padded rational base DWT"

What can i change to get it going?

Also not sure it was using the correct CUDA device when i tried some 3*2^n-1 exponents by hand as power consumption hardly got up when it ran as i have 2 CUDA devices in this machine. One 10 core card and one Tesla with 448 cores.

Help and user experiences with LLR cuda much appreciated!

Thanks,
Vincent

msft 2013-10-02 12:12

Hi,
69 is too large.


All times are UTC. The time now is 03:22.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.