Questions and Answers :
Wish list :
Better processing power management
Message board moderation
Author | Message |
---|---|
Danny Kim Send message Joined: 8 Oct 01 Posts: 5 Credit: 107,414 RAC: 0 |
I've used SETI@home Classic for years, but I have not had a single instance of the computer lagging or slowing down or showing signs of CPU (processing) power mismanagement. BOINC, on the other hand, seems to show this blatantly. For example, Classic never had to be turned off even if I had to run programs that require a lot of system resource. Somehow, however, if I did the same with BOINC, my computer (3.4G HT) would get noticeably "choppy" even when I'm doing such everyday stuff as surfing or document editing. And I can only consider it a product of bad programming. I really hope that the new version would be designed so as to better manage system resources. |
Pascal, K G Send message Joined: 3 Apr 99 Posts: 2343 Credit: 150,491 RAC: 0 |
Hmm, I have a p4 3.0 and do not have those problems , you might want to look your system over, you may need to upgrade your drivers as I do....... Semper Eadem So long Paul, it has been a hell of a ride. Park your ego's, fire up the computers, Science YES, Credits No. |
Keck_Komputers Send message Joined: 4 Jul 99 Posts: 1575 Credit: 4,152,111 RAC: 1 |
This is likely more of a HT issue than a BOINC issue. The OS sees two CPUs and schedules accordingly. Since there is really only one CPU sometimes a process can get more than it's fair share. BOINC WIKI BOINCing since 2002/12/8 |
Danny Kim Send message Joined: 8 Oct 01 Posts: 5 Credit: 107,414 RAC: 0 |
> you may need to upgrade your drivers as I do....... A fair guess, but I would bet against that since I never use the graphics window. Even if I did, mine is up-to-date all around. |
Danny Kim Send message Joined: 8 Oct 01 Posts: 5 Credit: 107,414 RAC: 0 |
> This is likely more of a HT issue than a BOINC issue. The OS sees two CPUs and > schedules accordingly. Since there is really only one CPU sometimes a process > can get more than it's fair share. > I understand where you're coming from, but it makes no sense for a program of low priority (BOINC) to stall those of high/normal priority. (IE, games, etc.) Again, this was never the issue with Classic. And about getting more than a fair share, it doesn't. If I ran just one instance instead of two, it would never take more than 50% of my processing power. Even if it did, since HT technology was released before BOINC, it's BOINC's maladjustment to HT, not the other way around. |
Keck_Komputers Send message Joined: 4 Jul 99 Posts: 1575 Credit: 4,152,111 RAC: 1 |
> > This is likely more of a HT issue than a BOINC issue. The OS sees two > CPUs and > > schedules accordingly. Since there is really only one CPU sometimes a > process > > can get more than it's fair share. > > > > I understand where you're coming from, but it makes no sense for a program of > low priority (BOINC) to stall those of high/normal priority. (IE, games, etc.) > Again, this was never the issue with Classic. > > And about getting more than a fair share, it doesn't. If I ran just one > instance instead of two, it would never take more than 50% of my processing > power. Even if it did, since HT technology was released before BOINC, it's > BOINC's maladjustment to HT, not the other way around. > The way I understand how this problem occurs the working set of one or some of the processes is large enough to push everything else out of the CPU cache. This causes instructions and/or data to be moved to and from memory or even disk. The OS does not realize that this is occuring, it assumes that both CPUs have access to the full cache. So it may be scheduleing 2 incompatible processes to run at the same time. This can also cause a low priority process to steal cycles from a high priority process because the OS thinks it has an idle processor. Intel actually warns that HT may cause performace problems in some situations, presumably this is what they are talking about. BOINC WIKI BOINCing since 2002/12/8 |
©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.