I have a dual boot machine. When I checks the stats on the seti page I noticed that they were different when running XP, or SuSE. The machine is mostly idle. Is anyone else running setiathome? Any thoughts on boosting perfromance? Running under XP GenuineIntel Intel(R) Celeron(R) CPU 2.93GHz Microsoft Windows XP SP2, (05.01.2600.00) Measured floating point speed 1390.1 million ops/sec Measured integer speed 2412.94 million ops/sec Average upload rate 18.98 KB/sec Average download rate 4.31 KB/sec % of time BOINC client is running 81.8484 % Running under SuSE 9.3 with X GenuineIntel Intel(R) Celeron(R) CPU 2.93GHz Linux 2.4.21-297-default Measured floating point speed 733.63 million ops/sec Measured integer speed 1895.05 million ops/sec Average upload rate 1.09 KB/sec Average download rate 494.71 KB/sec % of time BOINC client is running 58.6482 % -- Collector of vintage computers http://www.ncf.ca/~ba600 Machines to trade http://www.ncf.ca/~ba600/trade.html Open Source Weekend http://www.osw.ca
Mike, On Saturday 19 November 2005 09:27, Mike wrote:
How similarly are the two BOINC clients configured? In particular, is one set to compute constantly and the other tied to screen-saver activation? That could easily explain the discrepancy. Another simple possibility is that you simply give the CPU more to do when you're running Linux than when you're running Windows. Randall Schulz
On November 19, 2005 1:19 pm, Randall R Schulz wrote:
They have the share the same configuration file. At first I thought that maybe it was computing the average including the time that the Other OS was running, but the number haven't changed much since I stopped running XP on the machine. I thought I had 9.3 on the machine but it is an upgrade 9.0 would a 2.6 kernel make that much difference? Just tried an upgrade but it yast can't read the packages from the SuSE dvd. -- Collector of vintage computers http://www.ncf.ca/~ba600 Machines to trade http://www.ncf.ca/~ba600/trade.html Open Source Weekend http://www.osw.ca
Hans du Plooy wrote:
Hans, Do not foret that, accordingly to berkeley.edu, the old seti@home client will cease to work on this December, 15. -- Rui Santos http://www.ruisantos.com/
On Mon, 2005-11-21 at 13:46 +0000, Rui Santos wrote:
Do not foret that, accordingly to berkeley.edu, the old seti@home client will cease to work on this December, 15.
Thanks, I noticed that when I used the client to check the performance of a new PC. I checked out the BOINC client, I'm not that impressed. You now require a username and password. I collaborated with a friend using his old (university, doesn't exist anymore) e-mail address, so now we can't continue.
On Mon, 28 Nov 2005 16:30:49 +0200 H du Plooy <hansdp-lists@sagacit.com> wrote:
As far as the (web) *server* is concerned, you can still log on with the account ID instead of with the e-mail address. I'm less certain about the (pc) *client*, but I believe that __all__ you have to do is to REPEAT whatever it was that you entered on the "change password" page at the web server. mikus
On November 19, 2005 12:27 pm, Mike wrote:
Just finished an upgraded to 9.3 and not the change I was expecting Measured floating point speed 676.59 million ops/sec Measured integer speed 2025 million ops/sec % of time BOINC client is running 59.1031 %
Mike, On Saturday 19 November 2005 19:55, Mike wrote:
What change did you expect and why did you expect any change at all? On how many runs were each of these reports based? How tightly did you control the mix of other processes running concurrently? Judging from those numbers, I'd suspect that they are within the range of normal variation and in fact reflect no real difference. Keep in mind, too, that cache utilization is a critical performance factor. Any other processes executing concurrently compete for cache use and degrade the performance of the BOINC client. The bottom line is that I don't think you can really infer any change based on those observations. Randall Schulz
Sat, 19 Nov 2005, by kenziem@sympatico.ca:
I've seen the same thing here. Even in wine the Windows client is faster than the Linux one.
Theo -- Theo v. Werkhoven Registered Linux user# 99872 http://counter.li.org ICBM 52 13 26N , 4 29 47E. + ICQ: 277217131 SUSE 9.2 + Jabber: muadib@jabber.xs4all.nl Kernel 2.6.8 + See headers for PGP/GPG info. Claimer: any email I receive will become my property. Disclaimers do not apply.
On Monday 21 November 2005 18:47, Linuxjim wrote:
Does anyone know if there is a gui BOINC client manager for KDE similar to KSetiSpy or Ksetiwatch?
KBoincSpy: <http://kboincspy.sourceforge.net/> -- Marcus Lund
Mike, On Saturday 19 November 2005 09:27, Mike wrote:
How similarly are the two BOINC clients configured? In particular, is one set to compute constantly and the other tied to screen-saver activation? That could easily explain the discrepancy. Another simple possibility is that you simply give the CPU more to do when you're running Linux than when you're running Windows. Randall Schulz
On November 19, 2005 1:19 pm, Randall R Schulz wrote:
They have the share the same configuration file. At first I thought that maybe it was computing the average including the time that the Other OS was running, but the number haven't changed much since I stopped running XP on the machine. I thought I had 9.3 on the machine but it is an upgrade 9.0 would a 2.6 kernel make that much difference? Just tried an upgrade but it yast can't read the packages from the SuSE dvd. -- Collector of vintage computers http://www.ncf.ca/~ba600 Machines to trade http://www.ncf.ca/~ba600/trade.html Open Source Weekend http://www.osw.ca
Hans du Plooy wrote:
Hans, Do not foret that, accordingly to berkeley.edu, the old seti@home client will cease to work on this December, 15. -- Rui Santos http://www.ruisantos.com/
On Mon, 2005-11-21 at 13:46 +0000, Rui Santos wrote:
Do not foret that, accordingly to berkeley.edu, the old seti@home client will cease to work on this December, 15.
Thanks, I noticed that when I used the client to check the performance of a new PC. I checked out the BOINC client, I'm not that impressed. You now require a username and password. I collaborated with a friend using his old (university, doesn't exist anymore) e-mail address, so now we can't continue.
On Mon, 28 Nov 2005 16:30:49 +0200 H du Plooy <hansdp-lists@sagacit.com> wrote:
As far as the (web) *server* is concerned, you can still log on with the account ID instead of with the e-mail address. I'm less certain about the (pc) *client*, but I believe that __all__ you have to do is to REPEAT whatever it was that you entered on the "change password" page at the web server. mikus
participants (9)
-
H du Plooy -
Hans du Plooy
-
Linuxjim -
Marcus Lund -
Mike -
mikus＠bga.com -
Randall R Schulz -
Rui Santos -
Theo v. Werkhoven