• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Back to bring the PAIN!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Holdolin,

now that you seem to have things running properly for a little while. theres one other thing you can do to push some more speed out of your system.

open your app_info.xml file, and on the 14th line you will see this:
Code:
<cmdline></cmdline>

add "-nobs" between the tags like this:
Code:
<cmdline>-nobs</cmdline>

save the file, and restart BOINC

this will make the app use "no block sync". it will use a full CPU core to feed the GPU, so you'll see increased CPU use (from about 20% to 50%), but the GPU jobs should speed up by a few seconds.
 
Well it's a greatly appreciated help to have the tips I've been given. Between an optimized client and better config settings my output is much higher than simply using the default client and settings. You guys rock out loud :)
 
Whoa....that's some serious output.

I'm fighting some CPU OC instability the past couple weeks. I need to dial that in.
 
Ok, I've got Ubuntu successfully installed and running Seti. Finally.

Let me know if there's any optimizations I should do, gsrcrxsi
 
If my math is right, at 35sec/unit average and 65 credits/unit average, I should be doing around 160k RAC. I guess we'll see once more units are validated.
 
with the current WU distribution (mostly BLC and only a little Arecibo), and factoring in a slight bump for CPU work, I'd say 150-160k yeah.
 
Yep, looks like 165k avg. so far. Should take out none1 this week, and Holdolin next week.

I have some more tweaking to do on the CPU front (low volt OC, tweaking number of cores on Seti CPU side), but it's doing really well. Pulling 400W on full load currently. Shutting down task gets on my other rigs because they're just 'uneconomic'.
 
I can't BELIEVE you guys are still around! Wasn't there like a zombie apocalypse or something! lol :beer:
 
Rufus, I noticed you’re running the CUDA 9 app. The All in one package is setup with CUDA 9 as the default, but it also includes a CUDA 10.1 app that should run considerably faster on your RTX card. What happens is that when you try to run a CUDA 9 (or lower) task on a CUDA 10+ card, it has to do a conversion to make it work, and wastes about 10 seconds at the beginning of each task before actually doing anything.

This will require you to upgrade to the nvidia 418 or newer drivers, however. The 410 drivers do not support CUDA 10.1, only CUDA 10.0.

After upgrading your drivers. You will need to edit your app_info.xml file with the name of the CUDA 10.1 app in place of the name of the CUDA 9.0 app. You only need to change 2 entries.
 
Getting an error trying to install the 418 or 430 drivers:

Code:
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 nvidia-driver-418 : Depends: xserver-xorg-video-nvidia-418 (= 418.56-0ubuntu0~gpu18.04.1) but it is not going to be installed
                     Depends: libnvidia-cfg1-418 (= 418.56-0ubuntu0~gpu18.04.1) but it is not going to be installed
E: Unable to correct problems, you have held broken packages.
 
I think I got it. I did the GUI updater to the 430 drivers, and editted the app_info file. It says it's running the 10.1 packets now.

Does it look right?
 
Back