|
|
Enhanced Optimized
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
OK!!! I think I have gotten somewhere!
I have 3 results now with one of my compiles that seems to be working!
this one has been validated: http://setiathome.berkeley.edu/resul...ltid=326284363
Waiting for this one to validate: http://setiathome.berkeley.edu/resul...ltid=325871709
This one has a SETI@Home Informational message -9 result_overflow
NOTE: The number of results detected exceeds the storage space allocated. error: http://setiathome.berkeley.edu/resul...ltid=326909117 but so does the same wu from a pentium machine! http://setiathome.berkeley.edu/resul...ltid=326909118 so I think it is actually a "noisy" wu.
I'm going to try another build this time before I jump the gun again and try to release it.
I would also like to say that I'm not actually modifying any code, I am just turning on optimization flags that seem to have been built in like -O3 -unroll loops, etc. Altivec is there too but I'm not sure if it is actually built into the code, so if it isn't then I bet Alex and Rick could make a screamer!
I found that the nightly build I was using seemed to be the problem with the last build, using the nightly tarball from 5-11-2006 seemed to fix that.
Now I am going to try to see if it will build against boinc version 5.5.0 instead of 5.3.12 (not sure if that will make a difference).
Then I will make a test release.
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
Thanks for all your efforts Boog, you're doing great! Sounds promising. I'm excited as well to see what Alex and Rick can wring-out of this this. Good luck w/ your testing and please let me know if I can help.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by Gecko_r7
Thanks for all your efforts Boog, you're doing great! Sounds promising. I'm excited as well to see what Alex and Rick can wring-out of this this. Good luck w/ your testing and please let me know if I can help.
Well, now I'm just starting to get frustrated, this last couple builds are actually taking longer than the original client!
eventually I may figure some of this out.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
As I Switched to the official app yesterday evening I can tell you it's not all your fault, they released WU very demanding since yesterday (claiming in the 9-10 hours on my G5).
Also there seems to be quite a lot of problems into processing WU if you look at the other crunchers on the same units as you, we're on a bumpy road…
Thumbs up Boog, thanks for all you efforts.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Aug 2005
Location: Golden Valley, AZ
Status:
Offline
|
|
I will see what it does on my eMac.
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
Please let me know how these do, and remember these "may" not work correctly, and are not as fast as the "old" version of seti. And I still bet Alex and Rick could make this faster than I can!
Thanks!
Boog[/QUOTE]
I'll give it a go on the G4.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
trying the G5 one, crunching happily for now (about 30% of a WU I had already in the making). It's noticeably faster (it'd better be!) and it crunches. Now will it validate?
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by lepetitmartien
trying the G5 one, crunching happily for now (about 30% of a WU I had already in the making). It's noticeably faster (it'd better be!) and it crunches. Now will it validate?
I sure hope it does!
All of the wu's I have crunched using this build have validated, even the ones that had the -9 error because the other computers that shared that wu had the same error.
http://setiathome.berkeley.edu/resul...ltid=326721031
http://setiathome.berkeley.edu/resul...ltid=327386606
here is the latest one that it just reported http://setiathome.berkeley.edu/resul...ltid=326957162 still waiting on it to be validated (as of 6am eastern).
I'm glad to see it is deffinately going faster! I have been messing with this so much I was starting to have trouble seeing if it is faster or if I was just wishing it is faster! lol
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Nov 2003
Status:
Offline
|
|
Will try it out on my Quad G5, Powerbook G4 1.5, G4 Mini, and eMac over the weekend.\
Crossing fingers...
Of course it "would" be nice to actually have Altivec optimizations included as well...
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by jedimstr
Will try it out on my Quad G5, Powerbook G4 1.5, G4 Mini, and eMac over the weekend.\
Crossing fingers...
Of course it "would" be nice to actually have Altivec optimizations included as well...
If I knew how to check to see if they are actually there I would. Altivec should be built into the ffwt libraries (and I compiled the g4 and g5 versions separetly and for each cpu specifically), then they are linked against the seti executeable.
So, at least that much is altivec, if there can be more done in the actual seti code then these workers could be awesome!
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Sep 2004
Location: London
Status:
Offline
|
|
Hi Boog,
I tried out the G5 version on my Quad and it seems to be working ok.
Here are the first 3 results
1/ here It initial validated where four others failed.
2/ here initial validated ok and was quickest at 11938.50 cpu time
3/ here fully Valid ok at 2nd quickest with 16991.76. Only just behind a time of 16304.64 for the quickest.
So well done. I notice that the wu's vary a lot in length of TTC (time to completion), don't know if that makes any difference to your optimization.
Maybe Rick and Alex could use this as a base for altivec enhancement, just a thought.
K.
edit 1246
(
Last edited by Knightrider; May 19, 2006 at 07:46 AM.
)
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
First 2 WU in the benchmarks
Here and there
I'm "2397079" (I'm not a number, I'm a free man!)
The "there" unit was partially crunched (about 50%) with the official app. I'm doing well for my old G5. If it goes that good later, we'll "only" need some altivec thrown in.
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
Another one for me
On the time gain, it did in about 5 hours a WU of "8 hours", note that this is from memory, so not very scientific. On the comparison side, I do roughly the same as non optimized 2.8-3 GHz pentium 4, but the same with I suppose crunch3r app for XP are faster still it seems (nearly twice).
So there's still room for improvement… (altivec… where is my altivec…)
On a side note, we are very very fast when crunching noise. Like here (note, it was with the first Bogg trial this one)
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
I'd suggest that a reference-WU be used (chosen by boog, I guess), and performance comparisons on different machines would be run against that WU (just like with the previous optimized workers). That way, we'd know for sure to what extent there are performance improvements in the optimized workers. This appears particularly useful, because the enhanced WUs seem to show much greater variability with regard to CPU time required to completion.
Just a thought...
Ron
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
Originally Posted by halimedia
I'd suggest that a reference-WU be used (chosen by boog, I guess), and performance comparisons on different machines would be run against that WU (just like with the previous optimized workers). That way, we'd know for sure to what extent there are performance improvements in the optimized workers. This appears particularly useful, because the enhanced WUs seem to show much greater variability with regard to CPU time required to completion.
Just a thought...
Ron
I agree. Between the WU length (AR) and CPU speed differences, it will be a real pain trying to measure and compare speeds. Need a reference unit.
|
|
|
|
|
|
|
|
|
Registered User
Join Date: Jul 2006
Status:
Offline
|
|
I started looking at the completion times for the WUs crunched with boog's app. The most interesting comparison was Gecko_r7's 4th WU: It's now fully validated and Gecko had the 3rd slowest time on his PM G4 MDD. The slowest time was from a PowerMac 10,1 (G4 mini)from some guy on Team AOL(!?).
Looking at the other WUs crunched with boog's app:
Knightrider's #2 (on a G5 Quad) has now fully validated and with the fastest time. Others were 2.8GHz & 3GHz P4s.
Knightrider's #3—2nd-fastest, just behind a Pentium M
Letpetitmartien—G5
“Here” WU was a short WU, now fully validated, and lpm was 2nd-fastest (by ~100 and ~200 seconds over 3rd and 4th place machines). Third- and fourth-place machines are Athlon 64-bit FX-60s running WinXP.
“There” WU was 2nd-fastest. Two slower machines were 2.4 and 3.2 P4s, which claimed same credit as lpm.
“Another One” now fully validated. lpm’s was the slowest time of three. The other machines were 3.2 and 3.0 P4s.
Of Gecko_r7 other 3 WUs, only #1 has been fully validated. Basically tied for slowest among 4 machines (other machine with similar time was Athlon 1.6Ghz running XP).
Note that none of the other machines appeared to have used optimized apps.
QS
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
Yep, this RA affair is confusing a lot things…
Got another WU coming… (the next after will be a lonnnnng one)
Also, you can name me LPM for short
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2002
Location: Kozani, Greece, EU
Status:
Offline
|
|
halimedia's idea is a great one. Any one of us could test the same WU twice, e.g. make a simple duplicate of your BOINC folder, test the speed using the official worker and then using boog's. That's what I'm up to right now, anyway... Will let them reach a certain percentage and let you know...
btw... Does boog's worker take advantage of bigfft_wisdoms?
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Sep 2004
Location: London
Status:
Offline
|
|
Boog,
Did you know there is a seti_enhanced 5.15 worker in beta testing for PC's and I suppose for Mac's as well.
K.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Nov 2003
Status:
Offline
|
|
Originally Posted by Knightrider
Boog,
Did you know there is a seti_enhanced 5.15 worker in beta testing for PC's and I suppose for Mac's as well.
K.
5.15 was released to the wild for Window's only... the beta versions 5.14 and 5.15 were all Windows only betas.... the only difference was trying to get the Win98 timer problem fixed. So for now Macs will remain 5.13 and Linux will stay 5.12.
As a sidenote, since Crunch3r's app 5.12 already instituted a timer fix, you can keep that version in your app_info.xml to handle all versions of Enhanced up to 5.15.
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Sep 2004
Location: London
Status:
Offline
|
|
Originally Posted by jedimstr
5.15 was released to the wild for Window's only... the beta versions 5.14 and 5.15 were all Windows only betas.... the only difference was trying to get the Win98 timer problem fixed. So for now Macs will remain 5.13 and Linux will stay 5.12.
As a sidenote, since Crunch3r's app 5.12 already instituted a timer fix, you can keep that version in your app_info.xml to handle all versions of Enhanced up to 5.15.
Thanks. I am predicting that BOINC will settle down fairly soon, so I was a bit alarmed. Still hopeing though.
K.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Nov 2003
Status:
Offline
|
|
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by Thanar
halimedia's idea is a great one. Any one of us could test the same WU twice, e.g. make a simple duplicate of your BOINC folder, test the speed using the official worker and then using boog's. That's what I'm up to right now, anyway... Will let them reach a certain percentage and let you know...
btw... Does boog's worker take advantage of bigfft_wisdoms?
I'm not sure about the bigfft_wisdoms.
I have tried to get the stock 5.13 version to run the test wu http://www.mkx.si/boinc/#verify
and it will not (my compile won't either)
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2002
Location: Kozani, Greece, EU
Status:
Offline
|
|
Just finished a quick benchmark between the stock 5.13 and boog's 5.13 versions. Seems like boog's version is just 3% faster... Could it be that it's just the same? Just for the records, I left the stock worker for half an hour and it reached 3.667137%. Next up was boog's worker on the same WU (different BOINC folder) for half an hour and it came up on 3.777613%.
Hope we'll have a bigfft_wisdom-enabled worker or an AltiVec one really soon...
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
The next WU, can't access the database so can't compare. Crunched in 4 hours instead of supposedly 6 (the official estimate is rather good), for what it's worth…
The trouble with these test WU is explanations on how-to are always atomic rocket science when it should be simple and straight (like a process included by the developers of the project in something well thought… sigh)
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Aug 2005
Location: Golden Valley, AZ
Status:
Offline
|
|
With Boog's worker I came in at 12:42:51 for my first unit.
Second one has just started.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by Thanar
Just finished a quick benchmark between the stock 5.13 and boog's 5.13 versions. Seems like boog's version is just 3% faster... Could it be that it's just the same? Just for the records, I left the stock worker for half an hour and it reached 3.667137%. Next up was boog's worker on the same WU (different BOINC folder) for half an hour and it came up on 3.777613%.
Hope we'll have a bigfft_wisdom-enabled worker or an AltiVec one really soon...
I wish I knew more, and I can't figure out why it seems the Mac versionof seti won't run the test work_unit.sah but my linux boxes will.
I get an error of:
SETI@home error -108 Unknown error
from boinc_init_diagnostics()
File: /Volumes/Sage/BOINC_Mac/seti_boinc_enhanced/client/main.cpp
Line: 225
and that is with the downloaded version from berk.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
What I can say to be clearer, the official app takes the time to more than the time (about +20%) that BOINC tells us so. Boog's takes less, the less depending of the WU so certainly RA related.
A side note on Altivec/wisdom files as there seems to be some confusion. Altivec is a hardware part of the G4 and G5 which is able to make some specific computations (like crunching in our case) faster, it's down to Berkeley not to be able to provide us from the beginning of SETI classic an Altivec app… so we have to do our own (or we are cutting both arms from our G4/G5 for this type of computation, just like non Altivec software in audio or video is just braindead, there are examples from the beginning of Altivec and editors have still not understood this). The wisdom file from what I've understood is a kind of library to help make FFT operations faster, a look up table (correct me if I'm wrong). So the optimized app in the end will be Altivec modified, it's what made the quad screams. And there's luck some wisdom file will be around I assume is the work that our fellows did on the now dead old app can be used.
There may be not a lot of optimization left (as the SETI developers say) in the windows versions, but there's quite some room of improvement on the OS X side, at least on G4/G5. The Intel macs are a different problem and best left for people more in the know on their issues in SETI.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by lepetitmartien
What I can say to be clearer, the official app takes the time to more than the time (about +20%) that BOINC tells us so. Boog's takes less, the less depending of the WU so certainly RA related.
A side note on Altivec/wisdom files as there seems to be some confusion. Altivec is a hardware part of the G4 and G5 which is able to make some specific computations (like crunching in our case) faster, it's down to Berkeley not to be able to provide us from the beginning of SETI classic an Altivec app… so we have to do our own (or we are cutting both arms from our G4/G5 for this type of computation, just like non Altivec software in audio or video is just braindead, there are examples from the beginning of Altivec and editors have still not understood this). The wisdom file from what I've understood is a kind of library to help make FFT operations faster, a look up table (correct me if I'm wrong). So the optimized app in the end will be Altivec modified, it's what made the quad screams. And there's luck some wisdom file will be around I assume is the work that our fellows did on the now dead old app can be used.
There may be not a lot of optimization left (as the SETI developers say) in the windows versions, but there's quite some room of improvement on the OS X side, at least on G4/G5. The Intel macs are a different problem and best left for people more in the know on their issues in SETI.
I may end up with another build in a day or 2, I'm trying some very aggressive optimization flags, I'll test it here at home for a while, and if it runs well then I'll let you all have a stab at it.
I'm also trying to add in the -faltivec compiler option to see if it will make a difference.
I really with I could test against the reference wu to see how much of a difference it will make, but it looks as if it is something on the mac side because I CAN get seti enhanced on my linux boxes to run the test wu
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
Ok, now we're getting somewhere. Maybe a different angle in looking at this.......
G4 1.33 5.4.9 Superbench client Boog 5.13 ap
http://setiathome.berkeley.edu/resul...ltid=328353443 AR=.418 37,786. secs
http://setiathome.berkeley.edu/resul...ltid=328349444 AR=.419 37,790. secs
Pentium M 1.6 Boinc 5.4.9 Crunch3r 5.13 ap
http://setiathome.berkeley.edu/resul...ltid=328202004 AR=.418 17,339. secs
http://setiathome.berkeley.edu/resul...ltid=328463393 AR=.417 17,258. secs
http://setiathome.berkeley.edu/resul...ltid=328458826 AR=.417 17,049. secs
http://setiathome.berkeley.edu/resul...ltid=325722560 AR=.418 17,307. secs
On S. Standard, My G4 ran 5.2.13 Superbench & Alpha5.3 avg WU was 5300 secs.
The Pentium M w/ Trux 5.3.12tx36 & Crunch3r sse2 ap for Pent M avg WU was 3300 secs
Basically, Pentium M was 60% faster per WU.
If comparing times of similar ARs is indeed the correct way to index same-length WUs, (I'm not totally convinced of this) then this comparison shows my M is now 118% faster in S. Enhanced, an @ +58% wider gap than before.
After looking at WUs using distributed 5.13 ap, vs. current edition of Boogs, I am seeing the @ same 3% speed improvement in similar AR units opt vs. non-opt as Thanar.
There must be much more untapped potential to still be realized optimizing for Altivec and maybe the FFTW implementations. If the Windows aps are already more heavily optimized out of the gate and don't have as much margin left for dramatic speed improvements as compared to S. Standard (per many thread comments on Seti board), it makes me believe there is potentially tremendous efficiency and speed gains remaining for PPC G4,G5.
Boog, keep it coming!
(
Last edited by Gecko_r7; May 19, 2006 at 11:51 PM.
)
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2006
Location: Paris, France, Europe, Earth, Sol
Status:
Offline
|
|
Originally Posted by boog
I may end up with another build in a day or 2, I'm trying some very aggressive optimization flags, I'll test it here at home for a while, and if it runs well then I'll let you all have a stab at it.
Yummy…
I'm also trying to add in the -faltivec compiler option to see if it will make a difference.
If it's coded with floating point operations it will make a real difference. If they coded with integers only (provided I use the right English term, I'm not used to maths in English) there's some heavy recoding then to do.
Whot? You didn't use the -faltivec already? You should be spanked and whipped! (joking)
I really with I could test against the reference wu to see how much of a difference it will make, but it looks as if it is something on the mac side because I CAN get seti enhanced on my linux boxes to run the test wu
Maybe if you elaborate a little our fellows here could be of some help? I've had headaches by reading the test WU blahblah but maybe there's some smart fellow with more wisdom (incidental pun) around.
Thanks for the effort.
2 new WUs
This one (still pending)
That one
So far it seems a full validation tour for now
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by lepetitmartien
Maybe if you elaborate a little our fellows here could be of some help? I've had headaches by reading the test WU blahblah but maybe there's some smart fellow with more wisdom (incidental pun) around.
I figured it out! you need a file named init_data.xml I saved it here: http://boog.is-a-geek.org/seti/init_data.xml.zip
then you can put the init_data.xml, seti worker, the reference_result_unit.sah and in the same folder rename refrence_work_unit.sah to work_unit.sah and run ./seti_boinc -standalone in a terminal (have to cd to the directory you put those files in and change the name to run to the filename you have).
|
|
|
|
|
|
|
|
|
Forum Regular
Join Date: Oct 2005
Location: Las Vegas, NV
Status:
Offline
|
|
Boog: Every WU processsed on my rig is validating. Not one problem with your current build.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by Gecko_r7
Boog: Every WU processsed on my rig is validating. Not one problem with your current build.
Sweet!
I am working on trying to get the right optimization flags for fftw to hopefully get a faster build, so for now I am playing with compiling and watching the test wu for a while.
Other than that my next build will be about the same, just hopefully faster!
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Aug 2005
Location: Golden Valley, AZ
Status:
Offline
|
|
We will be waiting to test it for you.
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
Originally Posted by boog
I'm not sure how wise it is to use the ref-wu available there. It originates from the 4.18 development source and was created in June 2004 (see file creation date). From what I understand, the wu-creation process has been changed from original 4.x SETI to enhanced. If there isn't a new ref-wu provided by SETI (in the enhanced dev source), you should probably pick our own. To do this, I'd suggest the following:
- make a backup copy of your boinc/projects/setiathome.berkeley.edu folder.
- crunch along (ideally with the standard worker, not your optimized one)
- check your results on-line until a new WU is reported and validated
- then identify that wu (by comparing file names) in the backup copy of the setiathome.berekeley.edu folder you made, isolate it and make it available for us to download
I'm guessing it would be best to pick a wu that shows an average cpu-time, not a particularly short one. The short ones may not be representative of the bulk of computations taking place during crunching of an average wu.
HTH,
Ron
(
Last edited by halimedia; May 21, 2006 at 06:49 AM.
)
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2002
Location: Kozani, Greece, EU
Status:
Offline
|
|
halimedia,
Downloading boog's WU and starting work on that could be a little tricky... It involves poking around with a handful of BOINC client's xml files. I wouldn't suggest such a strategy. In the bottom line, why should we undertake what you are suggesting? What are we trying to find out?
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
Originally Posted by Thanar
Downloading boog's WU and starting work on that could be a little tricky... It involves poking around with a handful of BOINC client's xml files. I wouldn't suggest such a strategy.
No, this is not the case. Running the seti worker for testing purposes can be done completely independently of boinc.
In the bottom line, why should we undertake what you are suggesting? What are we trying to find out?
We need a baseline to compare against. The enhanced WUs appear to be far more variable w/ regard to CPU time required than before, and ppl are testing on different hardware. If we want to find out how much of a performance increase an optimization delivers, we first need baseline data (of different machine architectures, ideally), and then test the optimized worker against that. Basic performance testing strategy...
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
NB: I'm currently running a 'self-made' test-wu on my quad just to see what goes on compared to the 'old days'. I'll let you know the results...
Cheers,
Ron
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by halimedia
I'm not sure how wise it is to use the ref-wu available there. It originates from the 4.18 development source and was created in June 2004 (see file creation date). From what I understand, the wu-creation process has been changed from original 4.x SETI to enhanced. If there isn't a new ref-wu provided by SETI (in the enhanced dev source), you should probably pick our own. To do this, I'd suggest the following:
- make a backup copy of your boinc/projects/setiathome.berkeley.edu folder.
- crunch along (ideally with the standard worker, not your optimized one)
- check your results on-line until a new WU is reported and validated
- then identify that wu (by comparing file names) in the backup copy of the setiathome.berekeley.edu folder you made, isolate it and make it available for us to download
I'm guessing it would be best to pick a wu that shows an average cpu-time, not a particularly short one. The short ones may not be representative of the bulk of computations taking place during crunching of an average wu.
HTH,
Ron
There is also a wu (I think it is the same one) that comes in the latest source for seti. And the difference between "old" seti and enhanced is the way it looks at the data. Enhanced is looking "deeper" into the wu, hence the longer crunch times.
I have ran the wu (above) with the standard seti app, now I just need to run a couple runs with various trys of my builds.
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
Originally Posted by boog
There is also a wu (I think it is the same one) that comes in the latest source for seti. And the difference between "old" seti and enhanced is the way it looks at the data. Enhanced is looking "deeper" into the wu, hence the longer crunch times.
It would be good to check if it's the same. The following comment from the main seti-site makes me think otherwise:
May 16, 2006
All of our splitters have been altered to create workunits for SETI@home enhanced. Thanks for your patience during this transition.
Can you compare the two ref-wus (i.e. the one you posted the link for in your prevous post and the one you mention is included with the enhanced source)? It would be worth checking, I'd say...
HTH,
Ron
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
Thanks for the info, boog!
The files are almost identical, but not completely. There are three differences in the analysis configuration parameters:
gauss_null_chi_sq_thresh: mkx.si = 2.225, is-a-geek.org = 2.100
pulse_thresh: mkx.si = 17.0, is-a-geek.org = 20.0
triplet_thresh: mkx.si = 7.75, is-a-geek.org = 8.50
If you look at the creation dates, you also see that the one from your URL (from the current SETI enhanced source code, I assume) has a much more recent creation date. I would suggest that only the newer version be used for testing. Any other opinions?
(
Last edited by halimedia; May 21, 2006 at 04:20 PM.
)
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by halimedia
Thanks for the info, boog!
The files are almost identical, but not completely. There are three differences in the analysis configuration parameters:
gauss_null_chi_sq_thresh: mkx.si = 2.225, is-a-geek.org = 2.100
pulse_thresh: mkx.si = 17.0, is-a-geek.org = 20.0
triplet_thresh: mkx.si = 7.75, is-a-geek.org = 8.50
If you look at the creation dates, you also see that the one from your URL (from the current SETI enhanced source code, I assume) has a much more recent creation date. I would suggest that only the newer version be used for testing. Any other opinions?
You have a better eye than mine when looking at that file! I totally missed those, but I see them now.
They must have changed something in the way it is processed, maybe for the deeper scan?
They are both from the same scan of the sky tho, date:Fri Mar 3 08:36:01 2000 time:2451606.85836
|
|
|
|
|
|
|
|
|
Dedicated MacNNer
Join Date: Oct 2005
Location: Switzerland
Status:
Offline
|
|
boog: is there further documentation regarding the ref-wu in the enhanced dev source? Any other files (such as a ref-result)? I'm planning to put together a package with all necessary info, instructions, etc. Would be ideal if I could include the whole spiel...
TIA,
Ron
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Jun 2006
Status:
Offline
|
|
Originally Posted by halimedia
boog: is there further documentation regarding the ref-wu in the enhanced dev source? Any other files (such as a ref-result)? I'm planning to put together a package with all necessary info, instructions, etc. Would be ideal if I could include the whole spiel...
TIA,
Ron
Well, I searched all over to get the instructions and just pieced it together.
the files you need are http://boog.is-a-geek.org/seti/seti_test.zip
Then you put the seti worker in that folder (or a new folder with those 3 files in there) rename refrence_work_unit.sah to work_unit.sah. Open terminal and type "cd " (with a space) and then drag the folder you want to work out of from finder to terminal, hit enter.
you can give yourself something to copy by typing "ls" in the terminal (it will list the files so you can copy/paste the name of the seti worker if it is long and your lazy like me).
then run the seti worker "./seti_boinc"
the options you can pass to it are
options:
-nographics run without graphics
-version show version info
-verbose print running status
-standalone
I havn't seen a difference in running with and without the -standalone option
I would have had more results to post, but when I got up this morning, I found that while I was trying to test a worker, my install of seti was running without boinc!! Not sure what happened, but I will be watching to see if it was something that I triggered or not.
Hope this info helps!
|
|
|
|
|
|
|
|
|
Mac Elite
Join Date: Apr 2000
Location: Minneapolis, MN USA
Status:
Offline
|
|
As of Saturday Morning at 11:30 I still had non-enhanced 400 work units in my queue
and sometime between Sunday night and Monday morning I ran out.
I've switched over to your new client and will let you know the results.
So far the menubar client is "guessing" that each block will be 1 hr 22 mins but that
seems wildly optimistic. More later. This is on a DP 2.5 with 2.5 GB ram.
|
|
|
|
|
|
|
|
|
Junior Member
Join Date: Feb 2001
Location: Austria, Europe
Status:
Offline
|
|
Originally Posted by Todd Madson
As of Saturday Morning at 11:30 I still had non-enhanced 400 work units in my queue
and sometime between Sunday night and Monday morning I ran out.
I've switched over to your new client and will let you know the results.
So far the menubar client is "guessing" that each block will be 1 hr 22 mins but that
seems wildly optimistic. More later. This is on a DP 2.5 with 2.5 GB ram.
Hi Todd,
I have the same machine and 13 WUs validated o.k. Time is in the 14,000-22,000 seconds frame, depending on what WUs you receive. Very frustrating!!! My P4 Dual-Core 3.0GHz Windows XP machine needs only 6,000-8,000 seconds and claims/receives the same credit! On the Mac I use the boog-app, on the P4 I use the Crunch3r-app.
Lots of room for improvements...
;-)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Forum Rules
|
|
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is Off
|
|
|
|
|
|
|
|
|
|
|
|