Forum > Linux

SETI MB CUDA for Linux

<< < (28/162) > >>

sunu:
pp, that's what I've been saying in all my posts to you, UPGRADE TO 185.18.14!!!

pp:
 :D

I'm old school. I don't blindly upgrade everything to the latest version without thought. I need to understand what has changed and how it will affect the rest of my system. I tried 185.13 right now but got computation errors on every WU. Back to 180.60 for now. I'll have a look at 185.18.14 after some more research...

pp:
185.19 also gives computation errors and it's the latest from nvidia's ftp. Guess my energy has to focus on the boinc version next... I'll just have to empty my queue before I continue experimenting.

sunu:
No. The latest version is 185.18.14, trust me. But as I wrote in my other post the right driver is just the beginning. There is much more for a proper seti install.

I'll try to write a guide for seti cuda in linux and post it on a new thread.

Richard Haselgrove:

--- Quote from: pp on 11 Jun 2009, 02:59:48 pm ---185.19 also gives computation errors and it's the latest from nvidia's ftp. Guess my energy has to focus on the boinc version next... I'll just have to empty my queue before I continue experimenting.

--- End quote ---

The actual 'discovery post' is in a development area not generally accessible, but the facts seem clear enough:


--- Quote from: sunu on 09 May 2009, 10:55:20 am ---Experiments with CUDA 2.2

I tried cuda 2.2 with 185.18.08 beta driver and our 64 bit linux app.

1. 185.18.08 driver isn't compatible with 2.1 cuda libraries. You have to install the 2.2 ones.
2. Current optimized app doesn't have any problem running with the new 2.2 libs.
3. With cuda 2.2, current linux app exhibits the same behavior with the windows app. It no longer uses a full core, only for the first few seconds and then cpu utilisation hovers around 0-2%. Now in linux we potentially (see #6 ) can also use a 4+1 config, not only 3+1 as it is now.
4. Computation time is better with 2.2. Using file creation/modification timestamps, as it is now impossible to get accurate computation times for 2.2 I got for a random 0.44 AR wu:
          2.1      9min 30sec
          2.2      8min 28sec
5. Results are strongly similar
6. While standalone it runs ok, under BOINC I couldn't make it run, I get instant computation errors.

--- End quote ---

So you need the v2.2 CUDA runtime and FFT library as well as the updated drivers. I don't know the answer to that point 6, though.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

Go to full version