Seti@Home optimized science apps and information
 
Welcome, Guest. Please login or register.
Did you miss your activation email?
09 Dec 2006, 05:10:31 am

Login with username, password and session length
 

 
 
Seti@Home optimized science apps and information  |  Optimized Seti@Home apps  |  Windows  |  Topic: Is it working? 0 Members and 1 Guest are viewing this topic. « previous next »
Pages: [1] Go Down Print
Author Topic: Is it working?  (Read 302 times)
Zaphod
Knave
*
Offline Offline

Posts: 5


View Profile
Is it working?
« on: 15 Nov 2006, 09:37:25 pm »

Installed (I think) the optimizer for amd dual core. Task now shows setiarhome enhanced 5.15.  If ths is the enhanced version how did it get on all of my machines? I have five on a basic network, only installed on one machine.

If it is working, and copied itself to all machines, how do I know it's ok for all machines? Two are dual core amd's, single core amd, and two hp intels.

PS for those of us over the age of 30 PLEASE CONSIDER USING A FONT LARGER THAN 6 POINTS SO WE CAN SEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE! Cheesy
Logged
BenHer
Pre-Release Tester
Knight Templar
*
Offline Offline

Posts: 358


View Profile
Re: Is it working?
« Reply #1 on: 15 Nov 2006, 11:05:16 pm »

Press <ctrl> key while looking at most browsers and use your scroll wheel to change font sizes.

The program only gets installed where you install it.  The EXE wont copy itself anywhere, you must perform that action.

Enhanced has been the "name of the normal" seti for many months now, this is just the optimized version of that code.

Which version you use on your varous machines depends on what they are able to do and how well each version works there.

Later AMD Dual core or single core (usually called Athlon 64 or Sempron or other AMD names) should use the generic SSE2 version.

Middle range AMD Athlons  (Athlon XP)  should use the Generic SSE version.

Lower end Athlons (Just 'Athlon'  with no XP) should use MMX version.

The HP Intels are, most likeley, at least Pentium 4 CPUs.  But you would need to check and confirm this.  The generation order of Intel naming is

Pentium III
Pentium 4
Pentium 4 HT (Prescott)
Pentium M
Pentium D
Core Duo
Core2 Duo

Any system between 4 HT and Pentium D would benefit from the SSE3 version, probably even the Core Duo (not core2).  Core2 would work, but not be as fast as KWSN version 1.3.
Logged
Zaphod
Knave
*
Offline Offline

Posts: 5


View Profile
Re: Is it working?
« Reply #2 on: 16 Nov 2006, 01:09:20 pm »

Otay,  I thought I followed the instructions.   Sse2, copy all files to the Seti project folder.  Restarted Boinc, even restarted machine, no change.  Suggestions?
Logged
BenHer
Pre-Release Tester
Knight Templar
*
Offline Offline

Posts: 358


View Profile
Re: Is it working?
« Reply #3 on: 16 Nov 2006, 02:21:32 pm »

The current, already started, WUs will finish computing with whatever cruncher was used to start crunching them.

The new cruncher shouldn't begin working until a new task begins crunching from the beginning, and with every newly started WU from there onward.

[newly started = status( ready to run) and progress( 0.000% ) ]
Logged
Zaphod
Knave
*
Offline Offline

Posts: 5


View Profile
Re: Is it working?
« Reply #4 on: 16 Nov 2006, 08:01:09 pm »

waiting
wairing
waiting for the worms
to come
Logged
Nebby
Knave
*
Offline Offline

Posts: 1


View Profile
Re: Is it working?
« Reply #5 on: 16 Nov 2006, 09:27:30 pm »

Ben,

It started using the new binary for me right away, even with WU's that were already started.  You can't miss the new executable in Task Manager.

Thanks to you and the rest of the gang for the improvements. 
Logged
Simon
Shrubbus Maximus
Lord o' the Board
Knight who says 'Ni!'
*****
Offline Offline

Posts: 519



View Profile WWW
Re: Is it working?
« Reply #6 on: 16 Nov 2006, 10:10:51 pm »

Nebby,

that's what usually happens - in fact, I'd say always, but it's not as apparent if you don't check using task manager.

BOINC never says anything else but "found app_info.xml; using anonymous platform".

Thing is, if you resume a WU that you started with a previous app, it will not show up in the stderr (and hence, look like that WU was crunched with the previous app in the BOINC stats). This is because Ben changed the sources so it only writes the stderr output when it's below 0.1% (I think, correct me if I'm wrong Ben Smiley) of WU completion. Reason for this being, when you suspend/resume a WU often, the stderr output would otherwise repeat a lot.

Hope that explains things,
Simon.
Logged
Josef W. Segur
Global Moderator
Knight o' the round Table
*****
Offline Offline

Posts: 158


View Profile
Re: Is it working?
« Reply #7 on: 16 Nov 2006, 11:09:53 pm »

...
This is because Ben changed the sources so it only writes the stderr output when it's below 0.1% (I think, correct me if I'm wrong Ben Smiley) of WU completion. Reason for this being, when you suspend/resume a WU often, the stderr output would otherwise repeat a lot.

It was based on percentage for awhile, but now the identification info is only output if there's been no signal analysis done; i.e. when it's a new task.

There is a good indicator when a WU has been finished by the new version, it ouputs the signal and flop counts on two lines (one for labels and the numbers below). The older versions had individual lines for all 5 values with the label followed by the number.

Old style:

Code:
Flopcounter: 243250770573

Spike   count:  0
Pulse   count:  0
Triplet count:  0
Gaussian count: 0

New style:

Code:
Spikes Pulses Triplets Gaussians Flops
   0      0       1        0     790637442701
                                                                            Joe
Logged
BenHer
Pre-Release Tester
Knight Templar
*
Offline Offline

Posts: 358


View Profile
Re: Is it working?
« Reply #8 on: 17 Nov 2006, 01:04:49 am »

Hey - I've even optimized the stderr output....five lines down two lines  Roll Eyes
Logged
Pages: [1] Go Up Print 
Seti@Home optimized science apps and information  |  Optimized Seti@Home apps  |  Windows  |  Topic: Is it working? « previous next »
Jump to:  


Quote!
One, a robot may not injure a human being, or through inaction, allow a human being to come to harm; Two, a robot must obey the orders given it by human beings except where such orders would conflict with the First Law; Three, a robot must protect its own existence as long as such protection does not conflict with the First or Second Laws.
- Isaac Asimov, Laws of Robotics from I. Robot, 1950

 
Site Statistics
Total Members:264
Total Posts:1911
Total Topics:139
Total Boards:6
Total Links:6
Downloads
Apps
Windows Rev 1.x18065
Windows Rev 2.x5941
Linux Rev 1.x3388
Subtotal:27394
Source packs:1605
Tool/WU packs:1925
Total:30924
Pages served
Today:885
Total:493352
(since 6/26/2006)

Latest Member:
zvonas
 
 
Seti@Home optimized science apps and information | Powered by Enigma 2.0 (RC1).
© 2003-2006, LSP Dev Team. All Rights Reserved.
Seti@Home optimized science apps and information Forums | Powered by SMF.
© 2001-2005, Lewis Media. All Rights Reserved.
Powered by MySQL Powered by PHP Valid XHTML 1.0! Valid CSS!