I\'m about to see my first coupled model finish (yipeee & crossed fingers!). On my old P4 it\'s run over 2600 hours which combined with several other projects, has taken close to a year of calendar time.

I\'ll figure out a way to keep crunching CPDN but meanwhile I\'m wondering about plans for doing 40- or 80-year models or other variations which would be shorter than these work units. Are those plans real? Or just being discussed (or did I dream it :) If real, any update as to possible time frame?
Thanks

---

You didn\'t dream it. And I also think it would be nice to have some info.
And a regular update on results.
btw, the latest generation models give me over 1800 hours for 50% on my P4 (which took about the same time as yours for the previous one).
Good luck.

---

Forum search Site search

The plans are virtual. Or vapourware. Or Some day.
They\'re in the \"pushdown stack\", but at the moment there are a lot of other items on top. At least the data to allow for it is being collected.

There\'s a new model being tested at the moment which has less HD read/write. But it uses more ram.
It\'s a lot faster, but (currently) needs computers with SSE2.
Which of these is \"good news\", and which is \"bad news\", is up to readers.

So many programs, so little time.

The fact that the project considers faster crunching to be more important than shorter WUs is good news, in my opinion. I\'m in DC for the long haul, not daily thrills and gratification.

Ignoring time lost to maintenance and repair, my computers crunch 24 hours a day, 365 days a year or roughly 8700 hours per year. It matters not one bit to me whether I crunch 8700 WUs that run 1 hour each, 10 that run 870 hours each or 1 that runs 8700 hours, if I receive equal credit in all scenarios. If the scientists can learn equally from all scenarios then it makes no difference to anyone. Just do regular backups on long WUs, that\'s all.

---

Second that, Dagorath.

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

It\'s a lot faster, but (currently) needs computers with SSE2.

Sorry for being off topic.May I ask what is SSE2?
Thanks
Chris.

An enhanced set of processor instuctions.
First came MMX, then SSE, SSE2, SSE3, and now SSE4.
Wiki
(See the bottom of the page for links to the other instructions.)

It means that you can leap tall buildings in a single bound, instead of several smaller bounds and a long rope ladder. Sort of. :)

most any computer that\'s fast enough to run CPDN is going to support at least SSE or if its Intel based will most certainly support SSE2. im not shure exactly when AMD\'s started supporting SSE2 but id think anything fast enough for CPDN would have this.

---

Thanks very much Les and clownius.
The Wiki link states that SSE2 is supported by the athlon 64 series and pentium 4 and above,so perhaps any comp. bought after 2003 may be ok.
Thanks again
Chris.

....as backing up workunits is no help if a system dies or -- what a concept, one upgrades to NEW hardware sometime during that year (since workunits don\'t xfer to new h/w) ....

Hi

I transferred to new hardware and the WU were perfectly happy.

BW

Tricia

We all know that the researchers in Oxford, the cpdn developers, the mods and all the crunchers appreciate the effort everybody is making, and we all also know that with such big workunits the effort required is serious.

Carl has been hoping to sort out a way to get 40,80 and 120-year unfinished parts of crashed models completed on other machines, taking them from the 40-year restart dumps. Since then, however, the researchers have asked for new types of model to be prepared, and as we know, there are only three developers to handle the workload....

When MikeMars gets back from kayaking in Indochina, we\'ll be looking at what needs to be added to the cpdn READMEs. I expect this will include a recent succinct post by Astro on how to transfer boinc + model to a different computer. If any of you experienced crunchers can in the meantime find a moment (!) to look through the READMEs to see if you notice any other gaps, your help would be appreciated.

And if you\'re not backing up the boinc folder from time to time, item #1 by Les in the Crashes README gives a quick method. Countless models have been rescued and completed by restoring backups.

---

Cpdn news

This thread is now way off topic, and is becoming hostile.
Any more posts like the last few, and I\'ll delete the thread.

I quite agree, Les, and it would be a shame because sTrey\'s original question was a perfectly valid one.

Mavau asked

And I also think it would be nice to have some info.
And a regular update on results

There\'s a Milestones thread started by Poohbear in this forum\'s Cafe where I keep posting updates on our collective progress from the cpdn front page. But the best place to find all the news and announcements in a single thread is on the php forum. Just click on my sig.

---

Cpdn news

Getting back to topic:
Difficulties are being experienced with the highly optimised models.
And current tests are being run with the short 40 year models from BBC runs.

It looks like being a while until the new models show up outside of beta.

I\'m really sorry to see what this project has come to. Threads like this are very likely to scare away more crunchers, as it doesn\'t seem like the people behind the project are really listening.
Why is it that the developers have to be contacted by PM, and don\'t visit this forum? Isn\'t this forum on the project\'s own website?
A lot of useful and constructive thoughts have been provided here, but seem to be ridiculed or at best ignored.

I know this thread has become very much off-topic, but I feel this has to be debated nonetheless. Move it to another thread if appropriate, but don\'t lock or delete it, as it will give the impression that you don\'t care.

Why is it that the developers have to be contacted by PM, and don\'t visit this forum? Isn\'t this forum on the project\'s own website?

First off, CPDN started on it\'s own without BOINC, so their original forums which people are encouraged to use, are the ones that are monitored more than these. These were added because BOINC is not only the software you run, but the online database that the project runs, so these forums are just an added (IMHO) nescience. Keeping up two forums is hard work, on a skeleton crew. Since they are more comfortable with their old forums, that\'s what they use the most.

So, if you look at things in perspective, they figured these forums would be watched / policed / maintained by the volunteers who are moderators, and they run the other ones.

---

Hm, I see a lot of posts have been deleted from this thread, including mine.

Another observation is that the moderators Les and mo.v are not running the project themselves, how can you answer peoples questions if you are not in the \"dirt\" yourself?

If the admins don\'t have the time to look into this forum then why not shut it down?

If the admins don\'t have the time to look into this forum then why not shut it down?

That\'s my suggestion as well. Look in my other thread about the two fora.

---

Grüße vom Sänger

I can tell you that Les and Mo.v do work directly with the team on many things. They have been part of the project for a long time, and are authorized by the team to answer questions.

I personally would not want their status. I applaud them for their work.

My message about the fora was almost exactly on target with what Carl wrote. \"The powers at be\" say these forums must remain, while they want the other ones, also. So, two/three sets (counting BBC) will almost always be. It\'s something that has not changed in the couple of years that BOINC has been around, and would be near impossible to change now.

Some people will not \"make another account\" on the other forums. You have to register there, also. It\'s redundant to be registering in multiple forums.

People who use other projects are used to these type of forums. If they were taken away, and people were forced to make two accounts, I think it would cause less people to join.

Moving messages between the forums is not fesible anymore. If it was done early on, maybe, but not after over 2 years of using both.

---

deleted

---

Arnaud