Pages: [1]
Guest

2005-08-18 14:47:30
last modified: 2006-06-04 10:31:02

If you want to crunch those WU's faster, try these clients.

http://www.marisan.nl/seti/index.htm

These clients have been compiled to run best with whatever processor instructions your CPU can support. They're amazing! My 2.8GHz Prescott P4/533 Mhz/1Mb L2 went from 3 hours per WU down to 1:20 hours! They work like a champ!
Lee Carre
 
BAM!ID: 41
Joined: 2006-04-19
Posts: 262
Credits: 299,581
World-rank: 394,914

2006-01-11 16:01:41

Crunch3r's apps are even faster

they can be found here http://www.guntec.de/Crunch3r/setix86.html
Want to search the BOINC Wiki, BOINCstats, or various BOINC forums from within firefox? Try the BOINC related Firefox Search Plugins
Guest

2006-01-28 19:37:13
last modified: 2006-01-28 22:48:41


Crunch3r's apps are even faster


I agree with you about optimised clients from crunch3Rr.

Computer L1 cache L2 cache Floating point Integer memory Time/unitUn-optimised Time/unitOptimised

P4 640 3.2ghz Prescott 16k 2mb 2079 2934 1gb 1 h 45m 40m
P3 1000mhzCoppermine 32k 256k 970 2486 500mb 7h 30m 3h 45m
P3 800mhzCoppermine 32k 256k 691 1221 191mb 8h 12m 3h 45m
P3 600mhzCoppermine 32k 256k 527 920 128mb 10h 15m 4h 30m
P2 350mhzDeschutes 32k 512k 280 489 128mb 16h 30m 9h 30m


Best Regards,
Chris S.
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-01-28 20:56:28
last modified: 2006-01-28 22:48:17


[quote]
Crunch3r's apps are even faster


I agree with you about optimised clients from crunch3Rr.

Computer L1 cache L2 cache Floating point Integer memory Time/unitUn-optimised Time/unitOptimised

P4 640 3.2ghz Prescott 16k 2mb 2079 2934 1gb 1 h 45m 40m
P3 1000mhzCoppermine 32k 256k 970 2486 500mb 7h 30m 3h 45m
P3 800mhzCoppermine 32k 256k 691 1221 191mb 8h 12m 3h 45m
P3 600mhzCoppermine 32k 256k 527 920 128mb 10h 15m 4h 30m
P2 350mhzDeschutes 32k 512k 280 489 128mb 16h 30m 9h 30m


Best Regards,
Chris S.
part of
Wenn jeder an sich denkt, wird an alle gedacht!
Lee Carre
 
BAM!ID: 41
Joined: 2006-04-19
Posts: 262
Credits: 299,581
World-rank: 394,914

2006-01-29 05:45:15

I don't think so. I've tried a Crunch3r's optimized client on my AMD64 and it was doing work very strange. The client reported time of 8 mins but the actual time that was past from the start was around 12 mins
Well, I suppose that depends what version of his app you're using, and what OS it's on.

As far as I know, i'm using the latest (25th december 2005 release) for windows, with no problems except for the occasional app crash, but that's only 1 in every 1000 or so. For windows at least they're defenately quite a bit faster than TMRs
Want to search the BOINC Wiki, BOINCstats, or various BOINC forums from within firefox? Try the BOINC related Firefox Search Plugins
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-01-29 06:59:40
last modified: 2006-01-29 07:08:09

Well, I suppose that depends what version of his app you're using, and what OS it's on.

As far as I know, i'm using the latest (25th december 2005 release) for windows, with no problems except for the occasional app crash, but that's only 1 in every 1000 or so. For windows at least they're defenately quite a bit faster than TMRs

My tests are also based on the latest version for the AMD64 with SSE3 for windows. So i think that is not the case
part of
Wenn jeder an sich denkt, wird an alle gedacht!
Lee Carre
 
BAM!ID: 41
Joined: 2006-04-19
Posts: 262
Credits: 299,581
World-rank: 394,914

2006-01-29 09:15:20

[quote]
My tests are also based on the latest version for the AMD64 with SSE3 for windows. So i think that is not the case
i've heard there's only a very small gain with using SSE3 over SSE2, so you might want to try the SSE2 version
Want to search the BOINC Wiki, BOINCstats, or various BOINC forums from within firefox? Try the BOINC related Firefox Search Plugins
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-01-29 09:48:00
last modified: 2006-01-29 10:56:42

[quote]

Ah, i'm using SSE2 on a P4 and SSE on P3's
that's probably the reason

part of
Wenn jeder an sich denkt, wird an alle gedacht!
Saenger
 
Tester - Translator
BAM!ID: 5
Joined: 2006-01-10
Posts: 1735
Credits: 221,151,993
World-rank: 6,229

2006-01-29 14:25:04

[quote]
So i'm still staying with the marsian one )
Grüße vom Sänger
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-01-29 14:34:58

[quote]

So you found ETI?

(It's marisan, not marsian
part of
Wenn jeder an sich denkt, wird an alle gedacht!
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-02-10 19:58:59

[quote]
Thanks for the tip
part of
Wenn jeder an sich denkt, wird an alle gedacht!
Guest

2006-03-09 01:53:23

I can't remeber which one i use but I think it's Marisan. I am running a 1.5 Ghz Celeron M processor with 1 Gb of ram. Without optimized client was around 2 hours 40 min. with is about 50 min.
Guest

2006-03-24 19:25:46
last modified: 2006-03-24 19:27:35


Thanks for the tip


The real time (stopped manually by clock) will always be a little more than pure cpu time.
Other threads/apps have to get some cpu time to do their work, like anti-virus, the BoincManager or BoincView, firewall, browser, e-mail clients, etc. etc.

Or do you say the reported time in the manager differs from the reported time on the result page at berkeley?
In that case I should ask: Do you use the calibrating core client by truXoft? His version manipulates the reported time and mflops for every wu to get a fair amount of claimed credit. So it is normal to get a different time on the result page than in the manager.


AndyK
Guest

2006-04-07 12:32:13
last modified: 2006-04-07 12:33:42

Seems weird. Because I'm using Crunch3r's Athlon 64 client with SSE3 instructions (given mine has a Venice core, which added SSE3 instructions to it), and have had no such timing issues. Getting about 44 minutes/WU...

Mine's also the 12/25/05 version...

Can't explain the discrepency in results, unless it's dependent on a difference in core steppings in our respective A64s, or some other difference in the respective systems... Hmm
Son Goku 3SSJ
 
Forum moderator - Translator
BAM!ID: 8
Joined: 2006-01-11
Posts: 182
Credits: 2,572,985
World-rank: 125,252

2006-04-07 18:34:36
last modified: 2006-04-07 18:35:08

[quote] Seems weird. Because I'm using Crunch3r's Athlon 64 client with SSE3 instructions (given mine has a Venice core, which added SSE3 instructions to it), and have had no such timing issues. Getting about 44 minutes/WU...

Mine's also the 12/25/05 version...

Can't explain the discrepency in results, unless it's dependent on a difference in core steppings in our respective A64s, or some other difference in the respective systems... Hmm
part of
Wenn jeder an sich denkt, wird an alle gedacht!
Pages: [1]

Index :: Optimized clients :: Optimized S@H clients
Reason: