Message boards :
Number crunching :
"Laggy" GUI and Processor Scheduling: Win32PrioritySeparation
Message board moderation
Previous · 1 · 2
Author | Message |
---|---|
cov_route Send message Joined: 13 Sep 12 Posts: 342 Credit: 10,270,618 RAC: 0 |
Mike would you have any idea if it is best to make sbs a a factor of 64? clinfo says max workgroup size 256, preferred work group size multiple 64. Is wg related to sbs? |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
I`ve tried a lot values already on my old 5850 and my new 7970. sbs 256 is always faster. On my sons 6850 sbs 192 is best. It seems it depends on motherboard OS/RAM combination. On a 7770 i had for testing -sbs 156 was faster. With each crime and every kindness we birth our future. |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
Thanks Sten. Typos. With each crime and every kindness we birth our future. |
TimK Send message Joined: 19 Sep 08 Posts: 10 Credit: 6,079,103 RAC: 0 |
Thanks Mike, I've tried that, but it doesn't seem to have had much/any effect. It's hard to tell, because as I was putting it into the app_info file the lag decreased quite a lot on its own, after about 2 weeks of quite bad lag. |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
For the past couple of weeks my machine has also been "laggy". I checked a few things in performance monitor and found about 20,000 average page faults per second, hardly any hard faults. I thought this was high, until I saw the graph below showing nearly 500,000 per second. Lags on Ati MB is AR dependent. For "nearly VLAR" cases increasing period iterations number could give positive effect. But not all cases covered by this option unfortunately. -sbs switch changes PulseFind kernel execution time too so it will not only help to optimize execution times but also has ability to decrease lags (so, trying settings Mike proposed could help im good number of cases). But I currently don't see how effects from this options (especially from iterations num) can correlate with page faults. For -sbs option such correlation could be possible indeed cause this option also changes total allocated GPU memory. If increase in page misses means memory swapping between OS memory and GPU memory by driver, this decrease in -sbs value could reduce page misses. Could someone explore this dependence ? (between -sbs value and frequency of page missings spikes). SETI apps news We're not gonna fight them. We're gonna transcend them. |
TimK Send message Joined: 19 Sep 08 Posts: 10 Credit: 6,079,103 RAC: 0 |
Well, I've found why my machine seemed very laggy. The mouse was dying! As well as the lag I started to get what seemed random disconnect-reconnect sounds for a day or so. Then, as well as this, the occasional message "the last usb device connected has failed" (or something similar) and the mouse froze until I unplugged and replugged into another socket. Now I've tried a new mouse and the lag has gone, along with all the boops and beeps and freezing mouse. |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.