View previous topic :: View next topic |
Author |
Message |
Sir Richard Biggerstaff Knight


Joined: 27 Nov 2005 Posts: 45 Location: Maine, USA
|
Posted: Tue Nov 29, 2005 11:23 pm Post subject: Which projects should I join? |
|
|
I'm running BOINC on three PC's -- two old and decrepit, one reasonable. Until last week I was running the old SETI@Home client. Now that I've upgraded to BOINC, I find I have choices. (We used to say in my software days, "User Choice = Bad".)
Might I invite some opinions, prosyletizations, and / or rants on the relative merits - and stability - of the various project options?
Thank you and Good Ni. _________________ Sir Richard Biggerstaff
A Dark and Stormy Knight |
|
Back to top |
|
 |
ToeBee KWSN Castellan


Joined: 30 Jun 2004 Posts: 601 Location: Manhattan, KS
|
Posted: Tue Nov 29, 2005 11:59 pm Post subject: |
|
|
It all comes down to what projects you feel are worthy and/or fun to process for. I personally like LHC@home because they are using BOINC to help build the world's largest particle accelerator. It will make particles go BOOM! It is a concrete goal that will be accomplished within the next few years. The problem with that project is that their work supply is very intermittent. Sometimes they have work for 2 weeks and then nothing for 2 weeks. They have to run a batch and then analyze the results before they can create more work.
ProteinPredictor@Home has smaller work units that run well on my oldest cruncher (P3-500). They usually have a pretty steady supply of work. Einstein@home seems to be one of the most stable projects from what I have seen. They are searching for "gravity waves" for those who are into quantum physics.
Then there is Climateprediction.net... Don't run that on anything but top of the line. Work units take WEEKS of CPU time and I have had problems with it pushing "questionable" RAM over the edge into the "non-functioning" realm just because it uses a lot of memory and seems to access it heavily. Not the project's fault but their client will lay the smack down on your hardware While they haven't been the most stable of projects it doesn't matter as much since the work units take so long. And that is about all the projects I have direct experience with. Well except for seti. I am currently doing a lot of seti mostly because our seti team is lagging a little and may eventually fall out of the top 10 which would make our founder somewhat unhappy I'm hoping with classic shutting down, some of our big producers will move to BOINC and help boost our production.
I know some people are concerned about whether or not the results of a project will be open for review as opposed to being proprietary and potentially making someone rich off of free CPU cycles. I think I have seen particular concern about this on the protein folding projects since the results could possibly lead to new medicines being developed which would undoubtedly cost an arm and a leg to buy at your local pharmacy even if you helped to develop it by donating CPU cycles.
For what its worth, here you can see the breakdown of what projects I am currently doing. Those pie charts are always in flux with me as LHC comes and goes and I play around with BOINC on various computers. Your BOINC user name appears to be different than your forum handle so I can't look you up but you can look yourself up on the team stats site.
But above all, every shrub is precious no matter which field it is from!
 |
|
Back to top |
|
 |
alpha_fruit Prince


Joined: 16 May 2005 Posts: 5805 Location: Western North Carolina, USA
|
|
Back to top |
|
 |
Ib Rasmussen Knight


Joined: 02 Jun 2005 Posts: 35 Location: Denmark
|
Posted: Wed Nov 30, 2005 12:20 pm Post subject: |
|
|
alpha_fruit wrote: | F@H is a very good choice. |
But is it BOINC? _________________ /Ib
MPFC scripts: www.ibras.dk |
|
Back to top |
|
 |
ToeBee KWSN Castellan


Joined: 30 Jun 2004 Posts: 601 Location: Manhattan, KS
|
Posted: Wed Nov 30, 2005 1:04 pm Post subject: |
|
|
Ib Rasmussen wrote: | alpha_fruit wrote: | F@H is a very good choice. |
But is it BOINC? |
Not yet but they ARE working on a BOINC-based client. Last I heard it was moving kind of slowly though.
But my question is why is foding being pimped in the explicitly BOINC section of our message boards? Say... anyone know what this button does? Seems to say something about "deleting" things...  |
|
Back to top |
|
 |
LanDroid Prince


Joined: 11 Jun 2002 Posts: 4476 Location: Cincinnati, OH U.S.
|
Posted: Thu Dec 01, 2005 1:06 pm Post subject: |
|
|
I suggest concentrating on the two biology related Boinc projects, Predictor and Rosetta. One reason is these complement the folding@home project that so many knights are involved with. All three are related to protein structures and folding. By crunching three similar programs, our team can help make a bigger contribution in this area.
[/pimping] |
|
Back to top |
|
 |
Sir Richard Biggerstaff Knight


Joined: 27 Nov 2005 Posts: 45 Location: Maine, USA
|
Posted: Thu Dec 01, 2005 7:03 pm Post subject: |
|
|
I have stopped taking work on the LHC project; the sixtrack application seems to take over my PCs and not give them up. Is anyone else experiencing symptoms with it?
Look me up in the stats as "Richard Biggerstaff" if you're curious. _________________ Sir Richard Biggerstaff
A Dark and Stormy Knight |
|
Back to top |
|
 |
Sir Richard Biggerstaff Knight


Joined: 27 Nov 2005 Posts: 45 Location: Maine, USA
|
Posted: Thu Dec 01, 2005 8:37 pm Post subject: |
|
|
I get that there are Folders and Boincers, and coming from the old SETI project, I'm a Boincer almost by default. Why isn't running Rosetta and/or Predictor as valuable as F@H?
I imagine all the teams are seeing a bump in membership from people like me, who started seeing messages that the old SETI project will stop issuing work soon. The 15th, I think. I was a solitary knight, alone in the shrubberies for years, until I saw the light with Boinc!
 _________________ Sir Richard Biggerstaff
A Dark and Stormy Knight |
|
Back to top |
|
 |
Michelle Moistened Bint

Joined: 28 Oct 2004 Posts: 10233 Location: At my desk
|
Posted: Thu Dec 01, 2005 9:27 pm Post subject: |
|
|
I think they are just as worthwhile as each other, Sir Richard.
There are many good projects out there to participate in and I think it really boils down to personal choice. I'm running 7 BOINC projects plus Find-a-Drug (ending soon) on this computer, and I'm running F@H and eon on my other computer. Most of the Looneys are involved in many different ones, and switch between them as the need arises - usually when we need a boost in a project's stats. Very competitive business this.
Some Looneys favour F@H over BOINC, hence all the Folding pimps.
 _________________ My brain hurts.
Jammy's Brain Donor.
[img]http://www.katrinashome.com/KWSN_Michelle_counter.php[/img] |
|
Back to top |
|
 |
ToeBee KWSN Castellan


Joined: 30 Jun 2004 Posts: 601 Location: Manhattan, KS
|
Posted: Fri Dec 02, 2005 2:35 am Post subject: |
|
|
Part of the reason I prefer BOINC is because of its multi-project features. From the little glimpses I have seen here and there it looks like BOINC hasn't gone over very well with a lot of folding users. Though I guess the early versions of BOINC didn't go over well with a lot of people which is unfortunate however now a lot of people are bashing it based on their experience from over a year ago even though it has changed a LOT since then and has improved greatly.
Sir Richard Biggerstaff wrote: |
I imagine all the teams are seeing a bump in membership from people like me, who started seeing messages that the old SETI project will stop issuing work soon. The 15th, I think. I was a solitary knight, alone in the shrubberies for years, until I saw the light with Boinc! |
You are correct. Just look at the Seti@home combined stats. Starting about november 24th (right after they sent out the email and started displaying the popup) you can see a pronounced increase in the number of users. And since seti@home is the largest BOINC project, the same can be seen in the combined BOINC stats
Quote: | I have stopped taking work on the LHC project; the sixtrack application seems to take over my PCs and not give them up. Is anyone else experiencing symptoms with it? |
Nope. I am attached to LHC on 5 computers using both windows XP and linux and have never seen a problem with it. I see you are running 2k and NT - maybe there is an issue with pre-XP versions of windows? I don't know. |
|
Back to top |
|
 |
Dagger Prince


Joined: 10 Mar 2004 Posts: 2918 Location: BC Canada
|
Posted: Fri Dec 02, 2005 3:25 am Post subject: |
|
|
I'm thinking something about the number of handles increasing. I can't quite remember what exactly. _________________ When in doubt, kick it until it works. |
|
Back to top |
|
 |
Apple Knight


Joined: 26 Feb 2005 Posts: 74 Location: Castle Anthrax
|
Posted: Fri Dec 02, 2005 6:10 pm Post subject: |
|
|
Quote: | Then there is Climateprediction.net... Don't run that on anything but top of the line. |
Ya I just switched to it and I'm looking at the time till complete- 66+ days  _________________
www.saikano-usa.com |
|
Back to top |
|
 |
Sir Richard Biggerstaff Knight


Joined: 27 Nov 2005 Posts: 45 Location: Maine, USA
|
Posted: Sat Dec 03, 2005 9:08 pm Post subject: |
|
|
Yes, I found the work/reward ration a little thin too, at 6600+ hours to complete...
 _________________ Sir Richard Biggerstaff
A Dark and Stormy Knight |
|
Back to top |
|
 |
ToeBee KWSN Castellan


Joined: 30 Jun 2004 Posts: 601 Location: Manhattan, KS
|
Posted: Sun Dec 04, 2005 4:46 am Post subject: |
|
|
You DO get credit along the way. Every <insert correct number here> time steps it sends a "trickle" that gives you credit. I don't remember the exact number but on my 2400+ I think it usually sent 1-2 trickles per day depending on resource share settings, etc. |
|
Back to top |
|
 |
Dagger Prince


Joined: 10 Mar 2004 Posts: 2918 Location: BC Canada
|
Posted: Sun Dec 04, 2005 1:38 pm Post subject: |
|
|
Back when i was running CPDN every trickle was ~ 75 cobblestones. And at about 3 to 4 a day (64 3200+) it adds up fairly quick. _________________ When in doubt, kick it until it works. |
|
Back to top |
|
 |
Ib Rasmussen Knight


Joined: 02 Jun 2005 Posts: 35 Location: Denmark
|
Posted: Mon Dec 05, 2005 3:42 am Post subject: |
|
|
Dagger wrote: | Back when i was running CPDN every trickle was ~ 75 cobblestones. |
At the moment a trickle (= 10802 timesteps) in a normal wu pays ~94.5 credits, while for a sulphur project wu, which takes about 60% longer per trickle, you get ~160.7 credits. The sulphur project also has 5 phases of 24 trickles, instead of 3 phases, so a completed wu pays 19,281.57 credits.
CPDN seems to be pentium optimized, so if you are credit conscious and have an AMD, you should probably shrub for another project. _________________ /Ib
MPFC scripts: www.ibras.dk |
|
Back to top |
|
 |
|