Life After Seti: Where will TeAm members go?

[Overkiller]

Hey all,

I've been thinking about this more and more as the hypothetical end date approaches.
I don't crunch Seti for myself but under the Taboo Flag so its not really a personal calling,
It's more like a team thing I'm participating in .

I have been looking at various other PRojects that the team is involved in and I have been leaning towards a few
due to a personal draw. Coincidentally these are also projects where Help is necessary and even a few pcs can make a huge difference.

My Prospective List of Projects (In Alphabeitcal order ):

Climate Prediction:

Pros:

-Sounds really neat from the overall science standpoint, and the fact that one can SEE their simulation when its done, that must be pretty amazing.
-Would be great to run on systems without access to internet
-AT needs help in this project to stop it from its continual slide

Cons:
- Each Simulation DOES take 45 Days, Intermediate "WUs" are sent out but i would hate to lose 40 days of work due to a hardware crash.

DPAD:

Pros:

-Great Real-World Implications (Transmutation of Radioactive waste into harmless materials)
-Every PC makes a big difference
-DanC has kept the team alive and deserves some aid
-Ran this before the client became unstable through 2 beta revisions

Cons:

-Not finalized client yet
-Runs are giving valuable scientific data but not up to the level that final version of client will

FAD:

Pros:

-Finding Cures for HIV & Cancer sounds great in my book
-Team is picking up steam it appears

Cons:

-Stats are a bit funky
-Don't really know how often/much attn. project gets in improvements


F@h:

Pros:

-Worthy Cause, also non-profit, with real-world, perhaps SOON, implications
-team needs help to get into top 20

Cons:

-Client appears to be rather difficult to set-up with all optimizations enabled
-constant connection required more or less (not really an issue outside of outages on my broadband)

SOB

Pros:

-Great Client, easy to setup, nice stats
-Ran this for a long time

Cons:

-P4 Bug :|
-Proving Math Theory vs. other Science?


These are a few projects i am looking to switch to after seti 1 ends....Are there any others i should look into?
I dont really know where i shoudl go to as all projects appear worthwhile in my book.
Can anyone perhaps guide me in one direction or the other?

[edit] If it helps...i have a P4 2.8 @ 3.2, 1700+ @ 2.1, and a 1.2 Athlon @ 1.2, maybe another odd pc here and there

This is definitely not a flame fest, it is just an exploratory question as to where TeAm members think they will go after Seti 1 ends along with me trying to figure out the same thing


Let the discussion begin

 

[kloostec]

Hey, what about United Devices/Grid.org?

 

[Overkiller]

I ran that for about 2 years a few years ago but i have been told its a FOR-Profit organization??

The above listed projects i believe are largely non-profit and are university run.

 

[DannyBoy]

The P4 Bug with SoB will be fixed shortly, almost definately before SETI ends, so you should remove that con

 

[Overkiller]

hehe! i wonder what my 2.8 @ 3.2 would do. i remember o/c it to 3.0 b4 the bug hit and i hit 740K cems/s

also the problem does lie in the fact that the SOB team has a healthy (albeit due to the ars gauntlet) lead over the only competition, TPR both in productiona nd daily rate.

I would also sort of want to help out a struggling team

 

[networkman]

Our D2OL team would love to see a few more members!

 

[Overkiller]

Hmm. actually appears to be very much like Find a Drug just with a much more powerful TeAm presence!

 

[Allio]

Originally posted by: Overkiller
hehe! i wonder what my 2.8 @ 3.2 would do. i remember o/c it to 3.0 b4 the bug hit and i hit 740K cems/s

also the problem does lie in the fact that the SOB team has a healthy (albeit due to the ars gauntlet) lead over the only competition, TPR both in productiona nd daily rate.

I would also sort of want to help out a struggling team

We're winning in production because TPR pulled nearly all of their resources off SOB to help out their SETI team, as they thought there was a threat of them not finishing in first place. They're rather bitter about losing their first place in a mature project, and word on the street is that they'll be back to SOB with a vengeance as SETI ends.

Only time will tell.

 

[kloostec]

Originally posted by: Overkiller
I ran that for about 2 years a few years ago but i have been told its a FOR-Profit organization??

The above listed projects i believe are largely non-profit and are university run.

From http://www.grid.org/projects/cancer/faq_results.htm

Q: Are you going to sell the results of this project to large pharmaceutical companies?
A: No. The results of this study are the intellectual property of the University of Oxford and the National Foundation for Cancer Research, who will make the scientific findings of this project available to the greater scientific community.

Whereas FAD seems to be looking to sell their results...

 

[Overkiller]

Ah, who knows maybe i confused FAD with UD

Allio..word in the ARs forums is that one person will switch his fleet to SOB after SEti and that one person alone will cut AT's lead in 1/2 :Q

Only time will tell!

 

[BlackMountainCow]

If u fear to lose 40 days of work in case your HD crashes when running ClimatePrediction.Net I can assure you: won't happen!!!

You can, at any point of time, copy the whole climate folder to a cd-r or another HD and then do a repair install of CPDN over that folder again (after u changed your hd) and it'll crunch on right at the point it had stoped before

So that's one con less for CPDN! Actually u don't have any cons left for CPDN then

 

[Overkiller]

lol, i noticed that too

But there is no benefit to an HT enabled P4?
Does Climate Prediction just ignore the 2 imaginary processors and view the HT P4 as 1 or will it only run 1 instance of Climate Preidction thereby taking up only 50%? That would be a Con then! (b/c HT is so neat )

Also how often are intermediate blocks reported? Or are there "phases" to each run where data is saved/reported to central server?

Finally, what is the "base" cpu that was used to get a 40 day trial run?

If the 40 days is based on a P4 3.0 Ghz then pity my 1.2 :Q

 

[BlackMountainCow]

As far as I know there's as much benefit to HT as other DCs have. If u want to, u can run 2 instances of CPDN at the same time. But I got to confess that my knowledge in that point is very small, as I don't have any HT machine around So u could also run CPDN and F@H at once if u wanted to.

The intermediate blocks (or "trickle") r send to the server on a regular basis I have no idea of (as i have dialup ). What I do is that I check the server connection once a day (takes 3 secs) and that is my trickle. So that's nothing to worry about. If u have a steady line 24/7 CPDN will trickle regulary on its own, but I don't know the steps.

My "base" CPU was an Athlon XP +2000 no OC, on a MSI KT266 Pro board with 512 MB RAM. So nothing extraordinary. CPDN only uses up to 100 MB of RAM (coded that way) so it won't kill your resources any more than other DC projects.

 

[Overkiller]

any way to "tweak it" to use more ram/use only idle resources? Or is it coded to always run idle/ etc?

I might as well put it on my soon to be built 1.2 athlon as it takes it 6 weeks :|

hehe!

 

[BlackMountainCow]

As far as I know there's no way of tweaking yet. But I as I mentioned already, it's still in development and I'm somewhat sure that they'll optimize it more and more as time goes on. It's like SETI. Idle cycles only.

If u have a HT P4 u can run 2 instances of it I guess. That would double your output then.

 

[Overkiller]

Not necessarily Black Mountain. I believe that it would just double the length of time per Model

Also i have been reading the forums and it appears that this is perhaps the most stringent DC program OUT there.

The tinyest of mis-calculations can render a week of work down the tubes (i.e. one floating point integer was miscalculated but the climate reacted normally for another week until finally the problem compounded to kill the planet..the client has to reset to the last stable timestamp which was a week ago :Q...pretty scary stuff at how stringest everything is. The OC has to be rock solid, stable b4 attempting this project!)

Will look into it some more

 

[BlackMountainCow]

Well, there's the saying that the wings of a butterfly in Brazil can change the weather in the U.S. for example.

So yes, I think that this DC project can't be as "lax" as others.

Btw,I found out the trickle step. Like every 10,000 time steps. One timestep = 30 minutes of a model year. So every 200 calculated days it wants to trickle but it hasn't got to! (like once a day for me) But u can finish a whole run without any trickle at all.

 

[Overkiller]

Neat, There is some really great information in the CPDN forums about the process and the science :Q

Time for a short rest and then off to build the 1.2 athlon system and try CPDN

 

[BlackMountainCow]

Originally posted by: Overkiller... and try CPDN

YEAH!!!

 

[bunker]

Heh, this thread went from "Where will TeAm members go?" to "Recruit Overkiller to your project!"



Myself, I'll be going back to F@H or Distributed Folding.

 

[Jeff7]

I run F@H on my main system, and SETI on my secondary system. I'll probably move the SETI one to F@H; I'm also building another PC, mainly as an Internet/Printer firewall/server - I guess I could have F@H on it too.

 

[Starrider]

Seti phase 2 (someone had to say it)

 

[gistech1978]

FYI
-you cannot run more than one CPDN client at a time.

-it sends results to the server every 11,000 timesteps or so.

-45 days... it took a 1.6 willy about 45 days. my 2600+ at home did a full run in about 22-25 days or so.
i was getting over 1.5 model years/day give or take, so 45 days is for some older hardware.

-another warning about CPDN (and one of the main reasons ive scaled back) is the HUGE amount of data written to the drive almost constantly.

 

[BadThad]

I've started on the FAD project lightly....just one PC so far until SETI is done. Anything to help the war against cancer is worthwhile IMHO.

 

[D347h]

ChessBrain:

Pros:

-Fun to help playing chess
-Working only for some time, good to run along with other DCs
-Small - fast to download and crunch WU's [about 30 seconds max each]

Cons:
-Need to be online to crunch
-Poor GUI [but they are making new version, pretty amazing !]