Welcome to the MacNN Forums.

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

You are here: MacNN Forums > Community > Team MacNN > Boinc 4.13 Clients

Boinc 4.13 Clients
Thread Tools
mikkyo
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 20, 2004, 09:11 PM
 
Up for beta testing are the boinc 4.13 clients.
Special Boinc Compiles (Beta Version)

There are 2 versions, one requires XCode1.5 to be installed(free from Apple Developer - just make an account).
The gcc3.5 compiles are up to double the whetstone speed of the regular gcc3.3 compiles.
The gcc3.5 compiles will also work on Tiger Preview 10.4.

No new Seti workers are needed, you can run the default or the optimized or altivec version you are already using.

Please give feedback if you encounter any issues.

Enjoy.
     
Drash
Fresh-Faced Recruit
Join Date: Aug 2001
Location: UK
Status: Offline
Reply With Quote
Oct 20, 2004, 10:07 PM
 
Got XCode 1.5 (standard install with CHUD 4 also, tho. I'm no power user, just a bit of light tinkering) but I don't seem to have gcc 3.5, just 3.3 and 3.1 (I think). Running the BOINC client I get the following error when I try to run benchmarks via SETI Control 3.0.1.

dyld: ./boinc_4.13_ppcG4 can't open library: /usr/lib/libstdc++.6.dylib (No such file or directory, errno = 2)

Is gcc 3.5 a separate download?

Regards
Ash.
     
mikkyo  (op)
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 20, 2004, 10:20 PM
 
I guess I'm mistaken then
I thought it was part of XCode 1.5, but it looks like it is only part of the 10.4 preview only.
Hrmmm.
You can't just move the library over either as you need a ton of other stuff.
Ah well.
The 3.3 version will work for everyone at least.
     
Drash
Fresh-Faced Recruit
Join Date: Aug 2001
Location: UK
Status: Offline
Reply With Quote
Oct 20, 2004, 11:01 PM
 
The gcc 3.3 build seems to be working fine - the benchmarks are once again higher than standard: eMac G4 1GHz 1GB 10.3.5

Before, using boinc_4.13_powerpc-apple-darwin:

Benchmark results:
Number of CPUs: 1
510 double precision MIPS (Whetstone) per CPU
1526 integer MIPS (Dhrystone) per CPU
Finished CPU benchmarks

After, using boinc_4.13_ppcG4:

Benchmark results:
Number of CPUs: 1
674 double precision MIPS (Whetstone) per CPU
1600 integer MIPS (Dhrystone) per CPU
Finished CPU benchmarks

Regards,
Ash.
     
spunky
Fresh-Faced Recruit
Join Date: Aug 2004
Location: Minnesota
Status: Offline
Reply With Quote
Oct 21, 2004, 03:10 AM
 
How could you find out what version of xcode ya got installed?, I got xcode i just forgot what version..... so any info would be helpful... thanks...
     
mikkyo  (op)
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 21, 2004, 04:20 AM
 
Get Info in the Finder on the XCode application in /Developer/Applications.
     
mikkyo  (op)
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 26, 2004, 04:04 AM
 
The boinc_4.13 optimized clients have been moved to the regular download page.
They are working and stable.
You can just replace your existing client with the one you download and leave your Boinc directory intact.
Just stop the client, put in the new one, and start that up and boinc will resume where you left off.

Combined frist-timer archives with the new client and seti worker will be available soon, as well as a newer dominator pack with a few more scripts too.
     
zac4mac
Mac Elite
Join Date: Oct 1999
Location: near Boulder, Colorado
Status: Offline
Reply With Quote
Oct 26, 2004, 05:38 AM
 
Mikkyo -

Nice jump in benchmarks, thanks for the optimization. I'll have to get this over to the other 2 boxes. Just started BOINC last week.

Before: DP - whetstone - 1073.1 MIPS
Int - dhrystone - 3432.9 MIPS

After: DP - whetstone - 1831 MIPS
Int - dhrystone - 3678 MIPS

G5 DP2.0/1.5GB/9800Pro/10.3.5
climateprediction.net

Z
     
zmaniac
Fresh-Faced Recruit
Join Date: Oct 2004
Status: Offline
Reply With Quote
Oct 27, 2004, 07:43 AM
 
Originally posted by mikkyo:
Up for beta testing are the boinc 4.13 clients.
The gcc3.5 compiles are up to double the whetstone speed of the regular gcc3.3 compiles.
Please give feedback if you encounter any issues.

Enjoy.
While the Dhrystone speed doubled, the Whetstones have been cut in half:

SHRDLU:~/Boinc4 jim$ ./boinc_4.06_ppcG5
2004-10-22 06:32:13 [---] Starting BOINC client version 4.06 for powerpc-apple-darwin
2004-10-22 06:33:14 [---] Benchmark results:
2004-10-22 06:33:14 [---] Number of CPUs: 2
2004-10-22 06:33:14 [---] 4819 double precision MIPS (Whetstone) per CPU
2004-10-22 06:33:14 [---] 5180 integer MIPS (Dhrystone) per CPU

SHRDLU:~/Boinc4 jim$ ./boinc_4.13_ppcG5
2004-10-22 06:35:25 [---] Starting BOINC client version 4.13 for powerpc-apple-darwin
2004-10-22 06:35:25 [---] Suspending computation and network activity - running CPU benchmarks
2004-10-22 06:36:26 [---] Benchmark results:
2004-10-22 06:36:26 [---] Number of CPUs: 2
2004-10-22 06:36:26 [---] 2200 double precision MIPS (Whetstone) per CPU
2004-10-22 06:36:26 [---] 11534 integer MIPS (Dhrystone) per CPU
This is a 2.5GHZ DP G5 on Tiger 10.4 beta.

I sure am looking forward to getting the new worker.

Thanks for the good work!

Jim
     
mikkyo  (op)
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 27, 2004, 12:18 PM
 
It looks like something else was running during the benchmark.
If you are crunching climateprediction, when the client tries to run benchmarks, the CP worker still starts up and fails to stop until a little bit into the benchmark.
To get a true benchmark, run the client in a separate directory with just your predictor or seti account file present, and take the best of 3 runs using -run_cpu_benchmarks flag.
Be sure nothing else it running, including top or process monitors (check then quit them).
     
zmaniac
Fresh-Faced Recruit
Join Date: Oct 2004
Status: Offline
Reply With Quote
Oct 27, 2004, 10:50 PM
 
Originally posted by mikkyo:
It looks like something else was running during the benchmark.
If you are crunching climateprediction, when the client tries to run benchmarks, the CP worker still starts up and fails to stop until a little bit into the benchmark.
To get a true benchmark, run the client in a separate directory with just your predictor or seti account file present, and take the best of 3 runs using -run_cpu_benchmarks flag.
Be sure nothing else it running, including top or process monitors (check then quit them).
No, nothing else is running during benchmarks. Three fresh runs and a generic PPC run for comparison:

SHRDLU:~/Boinc4 jim$ ./boinc_4.13_ppcG5 -run_cpu_benchmarks
2004-10-27 19:33:02 [---] Starting BOINC client version 4.13 for powerpc-apple-darwin
2004-10-27 19:34:03 [---] Benchmark results:
2004-10-27 19:34:03 [---] Number of CPUs: 2
2004-10-27 19:34:03 [---] 2199 double precision MIPS (Whetstone) per CPU
2004-10-27 19:34:03 [---] 10266 integer MIPS (Dhrystone) per CPU
2004-10-27 19:35:13 [---] Benchmark results:
2004-10-27 19:35:13 [---] Number of CPUs: 2
2004-10-27 19:35:13 [---] 2200 double precision MIPS (Whetstone) per CPU
2004-10-27 19:35:13 [---] 11597 integer MIPS (Dhrystone) per CPU
2004-10-27 19:36:25 [---] Benchmark results:
2004-10-27 19:36:25 [---] Number of CPUs: 2
2004-10-27 19:36:25 [---] 2200 double precision MIPS (Whetstone) per CPU
2004-10-27 19:36:25 [---] 11569 integer MIPS (Dhrystone) per CPU
2004-10-27 19:36:25 [---] Finished CPU benchmarks

SHRDLU:~/xboinc jim$ ./boinc_4.09_powerpc-apple-darwin -run_cpu_benchmarks
2004-10-27 19:39:30 [---] Running CPU benchmarks
2004-10-27 19:39:30 [---] Suspending computation and network activity - running CPU benchmarks
2004-10-27 19:40:31 [---] Benchmark results:
2004-10-27 19:40:31 [---] Number of CPUs: 2
2004-10-27 19:40:31 [---] 1334 double precision MIPS (Whetstone) per CPU
2004-10-27 19:40:31 [---] 4776 integer MIPS (Dhrystone) per CPU
2004-10-27 19:40:31 [---] Finished CPU benchmarks
The 4.09 results are interesting since they seem different than what I recall they were under 10.3, but I don't have a copy of those numbers and don't currently have a way to run 10.3.

Jim
     
wiraqocha
Fresh-Faced Recruit
Join Date: Oct 2004
Status: Offline
Reply With Quote
Oct 28, 2004, 07:01 AM
 
1.25 MDD G4 (7450) w/ 1.28G RAM:
4.11-
856/2024 856/2018 856/2015
4.13-
858/2024 858/2015 857/2022
( Last edited by wiraqocha; Oct 28, 2004 at 09:59 PM. )
     
mikkyo  (op)
Senior User
Join Date: Feb 2002
Location: Silly Valley, Ca
Status: Offline
Reply With Quote
Oct 28, 2004, 07:14 PM
 
There shouldn't be any significant change between 4.11 and 4.13 for benchmarking.
Did you run the benchmarks more than once and pick the best?
You will find with multiple runs the number vary a bit.
     
wiraqocha
Fresh-Faced Recruit
Join Date: Oct 2004
Status: Offline
Reply With Quote
Oct 28, 2004, 09:57 PM
 
Originally posted by mikkyo:
There shouldn't be any significant change between 4.11 and 4.13 for benchmarking.
After three tests with each ver .13 came out slightly ahead.
     
djohnson
Professional Poster
Join Date: Sep 2000
Location: Texas
Status: Offline
Reply With Quote
Oct 29, 2004, 11:23 AM
 
I am running the 4.13 client and testing SetiLogger X. WHOA!!! Much faster on than the previous versions!

26% complete in 40 minutes and 49 seconds.
     
   
 
Forum Links
Forum Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Top
Privacy Policy
All times are GMT -4. The time now is 04:30 PM.
All contents of these forums © 1995-2017 MacNN. All rights reserved.
Branding + Design: www.gesamtbild.com
vBulletin v.3.8.8 © 2000-2017, Jelsoft Enterprises Ltd.,