Optimised apps for MW are killing !
I have never had a problem with any other project.... run down the cache, stop Boinc client, copy app info xml and the app file to the project folder, then restart client and allow the project t get new work..the new app runs and the world is mellow.
But I have done the above so many times now, and keep ending up with the "intel x86 " 0.17 file downloading with the fresh batch of WU's, and then the regular crunching of 50 minute 0.17 WU's begins all over again...and they are "high priority..which doesn't allow me to finally use CUDA for some of these other faster cuda apps.
Can anyone pick what I may be doing wrong..?
thanksamillyun
veebee
I found despite the app-info XML file MW still uses the standard app.
I had to stop BOINC and delete the standard app.... only then did it see the optimised app as directed by the XML.
To get cuda working go to the lunatics web site and look in the forums for version 7 of the cuda optimised app.
From Milkyway News Archive: (http://milkyway.cs.rpi.edu/milkyway/all_news.php#137)
I've upated the alpha release of the v0.18 application. It should now compile correctly. It has some changes for a small checkpointing bug and more performance improvements. I'm going to let other people try it out for a bit before I update the stock application. So if all goes well stock should upgrade to 0.18 in a couple days. After we've updated the stock application to v0.18 and we see that it's working correctly and we have optimized versions of it available, I'm going to stop awarding credit to pre 0.18 compiled applications, because we don't have any other way to get people to upgrade their optimized applications. I'll post a news item as a warning and provide links to appropriate v0.18 applications so that everyone has time to upgrade their applications before I make the switch.
veebee are you having problems in Windows or Linux, and which BOINC version?
Windows, 32 bit I have gone back to 6.4.5 as I had nothing but trouble with 6.6.x.
As it is now, 5 MW WU's run at one time which is "ok", but if I run SETI, only one SETI WU will run and the climate model.
In the config file, have a feeling that you should explain to your boy that there is just one GPU and four CPUs.... and I think the file that does that is mentioned somewhere in this board..... brb
"<avg_ncpus>-1</avg_ncpus>
<max_ncpus>-1</max_ncpus>
<coproc>
<type>CUDA</type>
<count>1</count>" as per Rocoll, but I'd change the avg and max numbers for the cpus to 4 and THEN see how it runs.
QuoteI've updated the stock application for linux and os x to v0.18. The code is in the code release directory as mw_v0.18d. This has a couple new performance improvements (thanks to our users) which has it running 30% faster, or more. This should bring our stock application (and some of the other optimized versions of it) in line with the fastest optimized apps we're seeing out there. Because of this, in the next day or so we'll have another (much loved) credit modification. I'll most likely be dropping the credit rate by about 30-50% but at the same time removing the credit limit. Due to the nature of the performance improvements, older hosts should see better performance improvements than newer ones, so for most people the credit/hr should be the same, while faster hosts might lose some credit or gain some credit depending on where they are compared to the speed limit.
any idea on what their "speed limit" gauge is set to? :)
Quote from: Tixx on February 16, 2009, 04:47:40 PM
any idea on what their "speed limit" gauge is set to? :)
108 credits per core per hour
yeah but they are removing the cap, so they must benchmark it to a ghz amount instead..
for example they would benchmark 108 credits done by a single core in 1hr at 2.1ghz speed
then anyone at higher clock speeds should be able to clock more credits per hour?
the whole milky credit juggling thing i still havent figured out :)
I understand your confusion. Unless I have misunderstood then I think Travis has made a large miscalculation there or he is hoping that nobody says anything. If he drops the speed limit and reduces the credit by 50% to 15 credits per task, then a fast Windows host that takes for example 5 minutes will score, cough, cough, very highly. :wink
Still that's nothing because the GPU client on a HD4870 takes only 9 seconds, so if WUs are worth 15 credits then the cr/hr will be 6,000 until the daily WU limit is reached after 2.5 hours. :coolgleam:
I have a HD3850 GPU ready to take advantage of any temporary bonus/error such as this but my computer refuses to operate. They renewed the electricity cable to my house on Saturday and I powered down from constant PrimeGrid crunching. When I tried to restart an hour later fans repeatedly spin for 5 seconds, stop for 5 seconds, but no banana. Computer is kaput. :cry2:
I have done some troubleshooting and it is either the motherboard or the CPU gone to electrical heaven. Most disheartening. When I feel better about it tomorrow I will try my old Pentium D 935 and see what happens.
Currently I am soldiering on with my old A64 3200 doing some MilkyWay, but the days for easy MilkyWay credits for these old boat anchors are just about to finish I think. I'm more disappointed that my raid on PrimeGrid has been so horribly terminated.
Quote from: kashi on February 16, 2009, 09:56:54 PM
When I tried to restart an hour later fans repeatedly spin for 5 seconds, stop for 5 seconds, but no banana. Computer is kaput. :cry2:
You tried disconnecting the power cable, leaving it out for 5 seconds or so, then plug in and power up again? Sometimes that sorts out the problem.
But yes, the Windows apps are faster than even 64bit Linux, so I might have to see how I can get hold of some Windows licences if that's the case! The X3350 at 3.4 GHz could probably do them in 3 minutes too iun Windows, so that's 80 WU an hour or (at 15 credits per WU) 1200 an hour / 28,800 a day on the one quad. Hmmm.
Thanks for the suggestion, I'll give it a try tomorrow.
I really hope it's not the CPU, the place I bought it from said they had stock, took my money and then didn't supply for a month. I particularly don't want to have anything to do with them ever again.
Back on topic a bit, yes if Travis wants to restrict fast Windows machines to 108 cr/hr or less he will have to reduce the credit to 5.4, not 15. Owners of slower machines and possibly Linux users will not be too happy about that. He'll possibly choose to do it in stages to cushion the shock for many MilkyWay contributors.
Ah well I can't try the GPU app yet anyway without 64-bit Windows, so I won't really be missing anything there. It's just the thought of doing a WU in 25 seconds or so excited my imagination. Hmm, that's 15K credit (1,000 WU daily limit) in 7 hours at 15 per WU.
New optimised Linux v18d apps from speedimic. :)
http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=566&nowrap=true#11020
Excuse double post, but I thought this deserved it.
Just ran a couple stock 0.19 milky wus..
on a duocore 2.7ghz they ran just under 22min..
so its almost double improvement on the old stock app.
however, hopefully we'll get the new optimized app out to bring them down to 4min again :)
Windows 0.19 optimised apps are up (http://zslip.com/)
Interesting that there is now an optimised SSE4.2 app. Core i7 owners may be pleased that Cluster Physik/Gipsel is thinking of them.
Now that the 108 cr/hr limit has been removed I wish he would release a 32-bit Windows or preferably a Linux GPU app. My quad may have melted but my HD3850 could still shine.
Quote from: kashi on February 17, 2009, 07:17:27 PM
Interesting that there is now an optimised SSE4.2 app. Core i7 owners may be pleased that Cluster Physik/Gipsel is thinking of them.
I tried the SSE4.2 app on my i7 with no perceivable change to the completion times over the SSE4.1
It's the first time I've seen a SSE4.2 optimised app so they may still be getting used to using it.
I was wondering how much difference there would be.
It seems it sometimes varies according to operating systems/compilers used and perhaps even the skill of the optimiser. I remember with the original Linux 64 optimised apps by petros the SSE4.1 was a fair bit slower than the SSE3, but for the optimised Windows apps by Cluster Physik/Gipsel the SSE4.1 was the faster one.
Still it is good to have the extra instructions available, even if they are not used much yet. The potential is still there for the future.
ATI 38xx/48xx owners:
get in while its hot! I can do 4 WUs in 45s a2 12cr a piece on a lowly 3850/256. Yep 1cr/s
Noice!!
wish i had some ati cards at the mo'!
ive always been nvidia unfortunatly heh
on the same note, seti gpu cruchin pays crap! hahaha
Tixx.... check it out and you'll see it because the lowest cr is given, that is usually your wingman's score and he used a non-modified app
:hbang:
However..... the Astropulse with mod app is a very nice payer, if and when you get jobs that is.
actually yeah ive seen them pop up every now and then.
indeed they pay well, but they dont crunch on gpu, and take upto 30hrs to crunch away heh
That's without the modified app.
1137286634 4059139 28 Jan 2009 10:17:47 UTC 31 Jan 2009 8:48:08 UTC Over Success Done 211,263.90 779.86 772.71
1148803370 4719913 4 Feb 2009 1:28:38 UTC 4 Feb 2009 6:03:42 UTC Over Client error Aborted by user 0.00 0.00 ---
1149258051 3761023 4 Feb 2009 6:03:47 UTC 16 Feb 2009 7:34:39 UTC Over Success Done 416,873.70 772.71 772.71
1165165762 4568089 16 Feb 2009 7:34:50 UTC 17 Feb 2009 10:32:58 UTC Over Success Done 35,762.88 769.79 772.71
Guess which one is mine and which one was done with a normal app on an AMD Athlon(tm) 64 X2 Dual Core Processor 4200+ [x86 Family 15 Model 75 Stepping 2] or Intel(R) Core(TM)2 Extreme CPU Q6800 @ 2.93GHz [x86 Family 6 Model 15 Stepping 11] OR mine *grin* Intel(R) Core(TM)2 Duo CPU E6850 @ 3.00GHz [x86 Family 6 Model 15 Stepping 11]
QuoteSorry guys, but in order to have some semblance of cross project parity in this (at the request of other projects), we had to reduce credit again. New WUs should now be generating around 6.1 credit for stripes 79, 82 and 86, and around 9.2 credit for stripes 20 and 21. Other projects have been getting requests from their users to increase credit to match ours from their users, because they felt they couldn't compete -- and this is exactly the situation we'd like to avoid.
looks like my credit will be spiraling down from 108c/h to probably around 70~
need to get my hands on some ati's :)
At the previous rate of around 12 per WU it was already getting "uneconomical" to run Milkyway on my 64bit Linux quads. At 6.1 it's not looking good when compared to other projects with 64bit apps (like ABC, Cosmology and PrimeGrid).
Still good value on the 32bit Windows boxes though!
QuoteMore credit changes
February 18, 2009
I think the drop to 6/9 credits was a bit much (for now), so I've upped the credits to 8/12 credits for stripes 79/82/86/20/21. I'm really gonna let this sit for awhile (promise) and see how things pan out, unless something drastic happens.
Looks like installing the SSE4.1 app on iGnatious may be a good move.... at least to see if the credits/hr rise and maybe a windows one back onto TheGnat...... hhhhhhhmmmmm what one was it????
Well, installing modified app for V0.19 for XP (SSE3) (32 bit) has resulted in a rather steady cr rate of 8.33 per job, which has taken from 5 to 7.5 mins........ giving a return of between 66.64 to 99.96 given that the other core is working on either CPDN or WCG.
Am waiting to see what happens with the new app in iGnatious as am running down the old 'uns, which are paying *shudder* 6.14 crs for about 15-17 mins work (admittedly at 85% power due to...... ) So, will be interesting when I install the appropriate app (SSE 4.1 I think). They should be done by the time I wake up (and ABC would hopefully have rewarded me with another 1000 cr WU *grin* Hey... one can hope can't one??)
EDIT: Of course... then I gets a Seti AP that will pay 750 crs for 12-13 hrs work ..... maybe even less or 60 crs min per hr :mwalk:
From Mylkyway News Archive: (http://milkyway.cs.rpi.edu/milkyway/all_news.php#150)
The server won't be awarding credits to applications with a version number less than 0.16. If you're not receiving credit for your optimized application then you probably need to update it to a more recent version. There are links to valid optimized applications in our code discussion and number crunching forums. If any valid applications are having problems with this, please let me know.
Quote from: Furlozza on February 19, 2009, 03:17:54 AM
Am waiting to see what happens with the new app in iGnatious as am running down the old 'uns,
No need to run these down. The new app will process both the new and the old WUs fine (back to version 0.16) so you just need to stop BOINC, install the new app and XML file, restart BOINC and hey presto the WUs will resume from their checkpoint and continue, but just a lot faster. biggrin
GPU app for Windows 32-bit (and 64-bit) can be found here www.file-upload.net/download-1465057/Milkyway_0.19_ATI.zip.html.
Works only on recent ATI cards (3800 or 4800 series). Requires SSE2 capable CPU.
I don't know how to install it and it is still an experimental app so may give problems. However one user has it working successfully on his HD4850 and is very pleased to be receiving nearly a credit per second.
from this thread http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=589
Well I bit the bullet and installed the optimised app on three of my Windows boxen so I should make 500k tonight. I wonder how much difference I'll see in the coming days.
Quote
... very pleased to be receiving nearly a credit per second.
from this thread http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=589
That's insane... I'm excited if I get a credit a minute! Maybe I just excite easily biggrin
i installed the opt app's on 3 doze since v0.16 - 2 P4D's and one q9450
one P4D is getting about 1.6CR / min (v0.19)
and the other one + the quad are getting about 2 / min (0.16)
might roll back to v0.16 !!
my i7 is doing (with 0.19) MW WU's in times ranginf from 4 min 30 to about 9 mins.... so there's obviously a big cr/hr/core difference between those WU's....
well, several hours of crunching in both boxes....... and Wiki, you'll be happy to know that you should hit your desired credit/minute rate without any trouble in XP32 bit using just the SSE3 app (I think .could be running SSSE3 app, but can't remember and too lazy to go looking LOL)
Same applies with iGnatious and SSE4.1 app. The times per WU now have minor variances, depending on what the box says the 'time to completion' is in Boinc, but both are doing them in either <6 mins (~8 cr/unit) or even 10+/- mins (12+ cr/min)
Either way, it works out at about 60 - 84 credits per hour per core. The reason for the '-' is that not all units in a run are the same length, hence you may get a run of 5 mins then two 10 mins and so forth.
Have TheGnat happily doing one core on MW and the other on Seti AP and then will be switching the second core between CPDN and WCG (Gotta extractae digitae in that one.... cause MW is starting to leave it fur dead :cry2:) until next Seti AP arrives.
You're right oh Furry one, I'm cranking through them well now. Using the SSE4.1 on my E8400, and SSE3 on the others that don't support newer calls.
The E8400 is averaging 110 cr/hr/core. WUs take between 4 and 7 mins and return 8 - 12 credits.
The Xeon 5130 is averaging 60 cr/hr/core, which is up from about 22 average on the version I was running a couple of weeks ago.
So it was worthwhile after all!!
definitely taking iGnatious to the docs tomorrow. :hbang: 110 per core??? on an E8400?? I want some of that :rofl:
Ran my X3350 with Windows for a while. With 4 cores at 3.4GHz it can earn 517 credits/hour (based on 100 completed work units). That's just under 12.5k per day.
Unfortunately it's not as good under Linux.
Been having a run on primegrid for a while, and decided to go back to Milkyway.
On the i7 (not even really wound up yet - 2.8 Ghz ), using the opt app (for sse 4.2) Im getting 85.75 cr/ hr/ core (686 per hour - 16500 odd per day)
EVERY Linux computer I have, (tried sse3, ssse3 for 64 bit linux) is erroring out EVERY wu , .. before they even "process" anything..
I even detach MW to make sure it is a "clean" set up, ( then re-attach and make sure ther's no wu's download. .. shut boinc (client and manager ) down, replace the necessary files, restart boinc)..
same happens again.
To coin a phrase.."Why is it so ??" * spoken with a nasally, twangy, white wirey-haired crazy but cool scientist sort of a voice*
Are you changing the permissions of the mod apps to boinc?? (Top two)
Thanks Furlozza, +1 for reminding me ... AGAIN, what you told me last time I had this prob..
(BTW - anybody - when did this needing to change permissions start up ??)
No Probs, Veebs..... ya want 'dumb', read my post about TheGnat running three projects on two cpus at once :hbang:
As for the permissions..... well, it applies with 6.2.12 which is the standard download with 8.10 Ubuntu. When the files are downloaded, my user name has the right to do the top two thingies.... but Boinc apparently wants nothing to do with me.... or you, but just boinc.... narky little piece of work that it is at times.
In all likelihood (and will allow correction on this) I have a sneaking suspicion that it is a long standing thing with Ubuntu/linux as part of the internal control. When the stock files are downloaded from the projects, there is no need to muck about with permissions cause they are set within the project. Mod apps just don't follow "those rules".
I just cant remember ever having to change permissions before.... it seemed like when this new "split" system (files in C:\Programs files\.. & c;\Documents and settings\allusers\..) came in windows is when I started hearing of this in linux... anyways, they are running fine now, thanks.
The problem is that when you install via Synaptic it installs under user BOINC, but when you download a different version it downloads with your user permissions. If you then copy the files into what is a protected directory, you must change the ownership to BOINC or it is forbidden to run.
Bashhead ( doh doh doh doh doh .....) that should have been logical....
Some advice needed.
I run BOINC inside a 32bit Windows XP install on a host with an Intel E5335 CPU. I want to run the optimised app.
- CPU-Z reports the same thing when run inside the VM as on the host: Instructions supported include SSE3 and SSSE3.
- BOINC startup message reports the processor features include SSE and SSE2 but not SSE3 or SSSE3.
The Milkyway forum has no relevant info. So I'm trying the SSSE3 app to see what breaks. Anyone tried anything like this? Any ideas?
WW
Install the SSSE3. TheGnat's Boinc summary also didn't mention the SSE3 or SSSE3...... but CPU-Z did, therefore go fur it.
Yes, CPU-Z accurately reports processor capability, so use the SSSE3 app.
Thanks guys. I've run a WU through on the SSSE3 app and looking good so far on the computer results page - it was successful and took 75% less time. Just what I need for the coming AA :cheer1:
Dumb question time :hbang:
I use Kubuntu on my main machine and it takes about 1 second to change permissions for the new apps.
My other 2 Q6600 crunchers do nothing but run Ubuntu and Boinc.
I cannot for the life of me work out how to change the permissions to allow the new enhanced apps to work. :cry2:
Everything is so hard to find and hidden away in Ubuntu. I can't even access the Milky app folder - it says you don't own it and don't have permission to alter it ( or something very close to this).
In Kubuntu you just click on folder, open as root, change permissions to anybody to do anything, done.
Any help greatly appreciated to do this in Ubuntu.
teemac, in ubuntu, I start "Nautilus" as a super user from a console (sudo nautlius) - I think you use Konqueror?
go to the files in var/lib/boinc-client/projects/ "milkyway", and then right click on them then select "properties".. there is then a few tabs, one of which is "permissions". click on it and then
there are some boxes that have "you" as the "owner".... these need to be changed to "boinc-client" or "boinc".
Thanks mate - got it.
A lot more involved than in Kubuntu. but it seems to be working.
:thanks1: :thanks1: :thanks1:
If it makes you feel any better i broke a few Wu's working out permisions for optimised apps myself. All fixed now thankfully
You can also use Dolphin or Konqueror (install either through Synaptic). Then just launch them in a terminal using sudo .. and then as per VBs instructions. It only takes a few moments once you get the hang of it.
Because of the great Milky optimized app for 32 bit windows, I stopped using my Ubuntu installation :(
I really wanna go back to Ubuntu, but the lure of Milky credit in XP is tooooo hard to break.
Can someone please provide me, in a nutshell, how the 64bit Ubuntu optimized app and the 32 bit optimized XP app perform against each other?
ONYX I'm someone and I can provide an answer in a nutshell ....
Same computer (Q9450), same clock speed(3.2ghz), same type of work unit (ps_s22)
Ubuntu 64 - 16 minutes 5 seconds
Windows XP 32 - 12 minutes 45 seconds
dont venture to the dark side onyx!! stick to windoze ahahahh
Im running around 17-18mins on my Q9550 for each WU using optimised app for SSE4.1 on Kubuntu 64bit.
I actually ventured to the dark side for Milkyway. My Xeon X3350 (effectively the same as a Q9450) takes 33% less tim efor the same work in Windows as it does in 64bit Ubuntu.
It's getting approx. 142 credits per hour (on each core) when running Windows while I'm at work. But when I'm home it goes back to Ubuntu where it gets around 100 115 credits an hour (based on the last 20 work units). Both use the SSE4.1 app
My Q9450 gets about 107/hr & my Q9550 gets about 111. Both on Kubuntu. I'm more than happy with that as it's about 3 to 4 times the credit I get on most of my other projects ;D
On my XP 32bit Q9450 stock speed am av btw the low 500's - mid to high 500's secs
Just put Milkyway on my i7 last night and it is taking about 2 hours per WU I think this means I need to use one of these optimised apps. How does this work?