Message boards :
Number crunching :
Regarding CPU Usage
Message board moderation
Author | Message |
---|---|
Gosub Send message Joined: 20 Dec 99 Posts: 4 Credit: 911 RAC: 0 |
I didn't want to hijack another thread, so I figured I'd start a new on for this. In another thread, this was posted... "Under normal seti use, this is as it should be. If you are using 40%, seti uses the remaining 60%. If you're only using 1%, seti uses the remaining 99%. The CPU usage will always be 100%, though seti is only using however many % you are not using." Is this the way the Classic version works? Oh, wait... BOINC is running the Classic app in the background, isn't it? Or, is it? Ugh... I'm so confused. I've tried reading up on this but can't seem to find anything that explains in. -Stever |
Darren Send message Joined: 2 Jul 99 Posts: 259 Credit: 280,503 RAC: 0 |
> Is this the way the Classic version works? Oh, wait... BOINC is running the > Classic app in the background, isn't it? Or, is it? Ugh... I'm so confused. > I've tried reading up on this but can't seem to find anything that explains > in. This is the same way classic works, however the boinc version runs with even lower priority than classic. If you're familier with linux-type nicing, seti classic runs at nice 1, seti boinc runs at nice 19. It will yield the cpu to absolutely anything that asks for it, whereas classic would try to hold on to some of it. If you bring up a monitoring application that shows cpu usage by process, you should be able to watch as the seti usage immediately drops everytime you do something else with the computer, then moves back up when you shut down stuff. As far as the differences between the classic and the boinc version, I've never seen anything yet telling what the difference is. They're not interchangable, but I don't really think they're different either - just customized for stand-alone or incorporation in boinc. |
Heffed Send message Joined: 19 Mar 02 Posts: 1856 Credit: 40,736 RAC: 0 |
> As far as the differences between the classic and the boinc version, I've > never seen anything yet telling what the difference is. They're not > interchangable, but I don't really think they're different either - just > customized for stand-alone or incorporation in boinc. They're entirely different. |
Bullwye Send message Joined: 26 Apr 03 Posts: 5 Credit: 111,847 RAC: 0 |
Hmm I'm wondering HOW low the prority of BOINC really is... This night my two computers just crunchin a few WU's. One did fine but the other just did 2 WU... it is slower, but when i watched tsk manager the idel process took 13h... how can that be? |
Darren Send message Joined: 2 Jul 99 Posts: 259 Credit: 280,503 RAC: 0 |
> Hmm I'm wondering HOW low the prority of BOINC really is... > This night my two computers just crunchin a few WU's. One did fine but the > other just did 2 WU... it is slower, but when i watched tsk manager the idel > process took 13h... how can that be? In linux nice terms, the boinc software runs with a nice of 0, which makes it about average for the priority that most software runs at. At that priority, it's going to keep going, but the boinc software doesn't really do much so it's certainly not going to tie up your system. The seti software runs at 19, which is the lowest possible priority, meaning it will give up the cpu to anything else that asks for it. I don't know why you would have anything in the idle process though, unless you have it set to stop the seti program if you're using the computer (in your preference settings online). If it's set to work all the time, it should take all the idle time for itself. |
Bullwye Send message Joined: 26 Apr 03 Posts: 5 Credit: 111,847 RAC: 0 |
I have set it to work all the time and one machine is doing well but the other not, wich is strage because they both use the same settings set up online. and the programms runnig in background (like antivir, firewall etc.) are 99% the same as is the OS. Very strange... |
anarchy-inc.org Send message Joined: 8 Jul 02 Posts: 15 Credit: 172,935 RAC: 0 |
> Hmm I'm wondering HOW low the prority of BOINC really is... > This night my two computers just crunchin a few WU's. One did fine but the > other just did 2 WU... it is slower, but when i watched tsk manager the idel > process took 13h... how can that be? Just like in classic, every work unit is different. Some require more crunch time than others, you can't look anymore at number of WU's completed. Only the CPU hours cruncing that count. |
Bullwye Send message Joined: 26 Apr 03 Posts: 5 Credit: 111,847 RAC: 0 |
> Just like in classic, every work unit is different. Some require more crunch > time than others, you can't look anymore at number of WU's completed. Only the > CPU hours cruncing that count. > I think you missunderstood me... the 13h wasn't the crunching time of one WU, it was time recorded and related to the windows idle process.. in other words the machine did about 13h absolutely nothing... BUT Seti/BOINC was running, thats for sure... but all i know is that seti usues exactly this 'idle' time of the CPU so it doesn't disturb other processes that the user wants to run... MfG |
ror Send message Joined: 3 Jun 04 Posts: 28 Credit: 3,020 RAC: 0 |
> > Just like in classic, every work unit is different. Some require more > crunch > > time than others, you can't look anymore at number of WU's completed. > Only the > > CPU hours cruncing that count. > > > > I think you missunderstood me... the 13h wasn't the crunching time of one WU, > it was time recorded and related to the windows idle process.. in other words > the machine did about 13h absolutely nothing... BUT Seti/BOINC was running, > thats for sure... but all i know is that seti usues exactly this 'idle' time > of the CPU so it doesn't disturb other processes that the user wants to > run... > MfG > Check you've got it to work whilst you're using your computer. (otehrwise if something is using 5% of your CPU, seti won't use the other 95%). If you've got it set up right you should have 0seconds to system idle process. (If that's what you want rather) |
Bullwye Send message Joined: 26 Apr 03 Posts: 5 Credit: 111,847 RAC: 0 |
> Check you've got it to work whilst you're using your computer. (otehrwise if > something is using 5% of your CPU, seti won't use the other 95%). > > If you've got it set up right you should have 0seconds to system idle process. > (If that's what you want rather) > yes it is exactly what i want it to do. But about this 5% mark i don't understand you... i've seen that if i turn on my TV programm wich has between 20 and 30% CPU usage seti takes the rest... or did u mean something else?!? |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
> > Check you've got it to work whilst you're using your computer. (otehrwise > if > > something is using 5% of your CPU, seti won't use the other 95%). > > > > If you've got it set up right you should have 0seconds to system idle > process. > > (If that's what you want rather) > > > > yes it is exactly what i want it to do. But about this 5% mark i don't > understand you... i've seen that if i turn on my TV programm wich has between > 20 and 30% CPU usage seti takes the rest... or did u mean something else?!? > Check to make sure that there is not some other process taking all of the CPU. Check your settings to make sure that BOINC runs when you are doing something else. |
Darren Send message Joined: 2 Jul 99 Posts: 259 Credit: 280,503 RAC: 0 |
But if his system is showing idle process time, then something else is NOT taking all the cpu - there is cpu there that is not being used by anything, including seti. Bullwye, make sure that both of your systems are set to run boinc using your preferences, and make sure that they have both been issued an update preferences command. This will ensure that are in fact running as you've set it up in your account preferences online. Also, just to make sure, these are single processor without HT? If you have multi processors or HT and your account is set to use only 1 cpu it will still generate idle process on the other cpu. |
©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.