Completed this task after 8 resends. Can anyone beat that?

Completed this task after 8 resends. Can anyone beat that?

WOW ... thanks for finally put that to rest!

Think about what that did to the progression of that core task's stream. The project has set us goals of turning around WUs in 48 hours and that one took almost 14 days. So when we whine and go on about we want WUs to run faster and not crash I can only imagine what Nate's and the other researchers are thinking ... No *~!$**

There were a couple of 8800s and an 8600 rigs in that pile-up ... come on people, those cards will never finish a long WU in time, set your preferences to standard please. I understand that will be interpreted by some as being 'elistist' but really, it is not doing the project any good and your not getting points even if you don't crash the WU.

---

Thanks - Steve

Completed this task after 8 resends. Can anyone beat that?

WOW ... thanks for finally put that to rest!

Think about what that did to the progression of that core task's stream. The project has set us goals of turning around WUs in 48 hours and that one took almost 14 days. So when we whine and go on about we want WUs to run faster and not crash I can only imagine what Nate's and the other researchers are thinking ... No *~!$**

There were a couple of 8800s and an 8600 rigs in that pile-up ... come on people, those cards will never finish a long WU in time, set your preferences to standard please. I understand that will be interpreted by some as being 'elistist' but really, it is not doing the project any good and your not getting points even if you don't crash the WU.

Here are one user, who own slow gpu (gtx 260 216 SP). I run now http://www.gpugrid.net/workunit.php?wuid=3648678 wu. Elapsed time now: 17 hours 26 minutes. Only a little bit over an hour left.

There were a couple of 8800s and an 8600 rigs in that pile-up ... come on people, those cards will never finish a long WU in time, set your preferences to standard please. I understand that will be interpreted by some as being 'elistist' but really, it is not doing the project any good and your not getting points even if you don't crash the WU.

Here are one user, who own slow gpu (gtx 260 216 SP). I run now http://www.gpugrid.net/workunit.php?wuid=3648678 wu. Elapsed time now: 17 hours 26 minutes. Only a little bit over an hour left.

Show Crash is talking about long workunits crunched on a generation older card than your GPU. Your current workunit is a short one, so it's ok to crunch it on a GTX 260/216.

Completed this task after 8 resends. Can anyone beat that?

This is a quite 'lucky' workunit, because it had 6 errors and 2 timeouts. I have crunched a couple of *_7's, which had 7 errors before one of my hosts finally completed them. For example this one.

There were a couple of 8800s and an 8600 rigs in that pile-up ... come on people, those cards will never finish a long WU in time, set your preferences to standard please. I understand that will be interpreted by some as being 'elistist' but really, it is not doing the project any good and your not getting points even if you don't crash the WU.

Here are one user, who own slow gpu (gtx 260 216 SP). I run now http://www.gpugrid.net/workunit.php?wuid=3648678 wu. Elapsed time now: 17 hours 26 minutes. Only a little bit over an hour left.

Show Crash is talking about long workunits crunched on a generation older card than your GPU. Your current workunit is a short one, so it's ok to crunch it on a GTX 260/216.

Oops. Sorry. But now "short" wu is completed:
Runtime: 67,475.38 Sec CPU time 1,074.44 Sec Credits: 27,250.00

Just a couple observations on this "pile-up"...

The two biggest delays for this result were timeouts and one of those happened on a GTX 560 Ti. The other timeout was on an 8600 which I agree should not be crunching long tasks but I feel the tone of some of the comments is a little too harsh on slow card owners. Some volunteers, perhaps 50% or more, don't have a clue what model GPU they have, they have no idea how to find out what model they have and even if they did they don't have time or they just don't care. Therefore posting requests to those folks either here in the forums or even on the home page, even in big, bold, red text won't stop them from selecting long tasks for their slow cards. We all want that situation to be different, and we want every volunteer to RTFM not once but three times but let's face the facts... chickens will vote for Colonel Saunders before that ever happens.

My feelings on this issue is that if the admins are willing to send long tasks to slow pokes then obviously the resulting delays aren't a big problem for them. Otherwise let them determine the minimum model that can be issued long tasks and put an end to the problem. Maybe certain models should not be issued even short tasks. Sure, some folks will complain about elitism and being excluded and taking 1001 volunteers with them to other projects but those 1001 others never materialize. Bottom line is the admins have to decide whether their goal is to make everybody feel loved and wanted or to get the results back in timely fashion and get on with the research.

I used to wrestle steers at the rodeo. The first lesson I learned is that the boys who encourage steers to roll over finish last. The boys who take the animal by the horns and twist it down whether it feels encouraged to go down or not are the ones who have a shot at the money. So "encouraging" a 2 day return by allowing a 5 day deadline is dumber than a bag of smashed arseholes, IMHO. Might have taken a genius to program that but the idea itself is just plain stupid and I don't care whether that doesn't spread the happy, happy, joy, joy feeling or not. Want 2 day returns then make the deadline 2 days and be done with it. Confusing already apathetic and confused volunteers with more convoluted thinking has never worked in the past, it isn't working now and it ain't gonna work in the future. Not at this project, not in any endeavor.

I made a suggestion for a way older cards can contribute meaningfully to the project. I didn't need to start with 'come on people' and for that I apologize to any of the readers who found that offensive.

dumber than a bag of smashed arseholes, IMHO. Might have taken a genius to program that but the idea itself is just plain stupid and I don't care whether that doesn't spread the happy, happy, joy, joy feeling or not

I would like to suggest you take a look in the mirror before calling out other post's 'tone'

---

Thanks - Steve

Your first and second paragraph fits this thread and I agree with those, but the third...

I used to wrestle steers at the rodeo. The first lesson I learned is that the boys who encourage steers to roll over finish last. The boys who take the animal by the horns and twist it down whether it feels encouraged to go down or not are the ones who have a shot at the money. So "encouraging" a 2 day return by allowing a 5 day deadline is dumber than a bag of smashed arseholes, IMHO.

I disagree with you. I don't understand how is your rodeo parallel justify your humble opinion about the bonusing of fast retuns?
Using bad language won't help to convince me.

Might have taken a genius to program that but the idea itself is just plain stupid and I don't care whether that doesn't spread the happy, happy, joy, joy feeling or not. Want 2 day returns then make the deadline 2 days and be done with it. Confusing already apathetic and confused volunteers with more convoluted thinking has never worked in the past, it isn't working now and it ain't gonna work in the future. Not at this project, not in any endeavor.

You are confused here, indeed. They want 5 day returns, so the deadline is 5 days. If a cruncher wants bonus credits, he/she can have it by returning the result within 24 or 48 hours. This is a very effective way to encourage crunchers to have state of the art GPUs.