Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 17
Posts: 17   Pages: 2   [ Previous Page | 1 2 ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 2443 times and has 16 replies Next Thread
Speedy51
Veteran Cruncher
New Zealand
Joined: Nov 4, 2005
Post Count: 1220
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Project always loses data


My opinion is that if we had enough computers to complete their project in a year or so, they would probably remove the time limit.

Lawrence

Lawrence I may have misinterpreted what you mean when you say "they would probably remove the time". If they removed the time limit would that not cause the job to continuously run and never finish? Is that what you were meaning or are you meaning to remove the time limit to allow the job to run for as long as it takes to complete the 16 jobs within one task?
----------------------------------------

[May 28, 2013 10:19:25 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Project always loses data

Hi Speedy51,
I meant remove the 12-hour time limit that keeps CEP2 from finishing all 16 jobs in a CEP2 work unit on slow computers. I did not mean remove the instruction count limit that all projects have to bomb out if caught in an endless loop.

Lawrence
[May 29, 2013 1:05:52 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Project always loses data

You can do much better job letting your machine to crunch nonstop. I cant think of any better use for it, than adding its power to 10 PetaFLOP machine focused upon crunching CEP2 in a week instead of two years by 10% of total amount of host within WCG. Cray Inc hasnt approved CEP as first priority for "Titan". Thus, for two years we have been stagnating greatly.
We are not playing games here. What we compute has sense, because people on Earth drastically need alternative energy resourses, clean water and food in great amount to preserve ecosystem and human lives. After resolution we may think about rising theirs and ours living standards.
We dont have endless amount of time to solve all problems by means of supercomputers and distributed computing. Some say we have 10 years left until civilization will reach its peak (see "Home" by Luc Besson).
We will not be able to make widespread introduction of any product if product's cost will remain high. Within current economy problem resolves partially by effective altruism and production of electric energy within customer's private property.
Some people are poor because they have nothing to give to wealth ones. There are billions of them. The wealth ones and a billion of their servants are highly dependent on each other and several billions living worldwide to serve to one billion. Its not bad to serve or be served, cause its simple specialization, but why inequality?
Its a shame many people making money producing, for example, teddy bears and clay sculptures, playing music, sharing knowledge, producing some advanced electronics of DisplAir to avoid poverty, thus further approve participation in virtual development of civilization, while at the same time dozen of thousand of kids a day worldwide find themself dying having nothing to eat.
Solar cell is mass productive weapon of freedom, concerning commercial application or, more than that, free energy giveaway. Despite the price, its handsfree, what is useful when you need to build your own production lines, being in the same time sharing finances and property with village or town, making decisions about shopping. Sure, there is no go without altruism. No one has to be left behind, no one.
Summing up, society cant afford selfsustainable abundance without alternative energy sources, which are needed ASAP. Few Terawatts would be just fine for the start.
Thank you. See "Venus Project" for further details.
[May 29, 2013 6:59:13 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Project always loses data

As far as i can see population participation in scientific advancement by distributive computing has very low grade. I have my AMD Phenom 9650 (2GFLOPS) crunching CEP2 for 2.2 years of cpu time, My position is 34000+ in WCG global rating and advancing. I think i have rather obsolete CPU and im very surprised by my overall standing, bearing in mind there should be 10^4, 10^5 or 10^6 cpus built per year. But they are used for no more than for business and entertainment despite the speed.
Only a few million of hosts crunching out of billions worldwide. Smartphones present standalone 5 PFLOPS of computation power.
[May 29, 2013 9:42:48 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Speedy51
Veteran Cruncher
New Zealand
Joined: Nov 4, 2005
Post Count: 1220
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Project always loses data

Hi Speedy51,
I meant remove the 12-hour time limit that keeps CEP2 from finishing all 16 jobs in a CEP2 work unit on slow computers. I did not mean remove the instruction count limit that all projects have to bomb out if caught in an endless loop.

Lawrence

Hi Lawrence,
Thank you for the clarification. I was fully aware that you were only referring to CEP 2.
----------------------------------------

[May 29, 2013 9:40:39 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Project always loses data

Hello dkt,
The only question I have is why are the last four jobs even contained in the set of jobs for the WU? It seems to be it bit of a waste when it leads to questions like this being raised. If all that is needed/wanted are jobs 0 to 11, what is the point of having jobs 12 to 15?
The 16 jobs are arranged in groups in order of importance. The scientists want to cover the entire molecule space in as few years as possible. As long as the first 3 jobs are run, that is useful info, but they would like it all. However, careful thinking on their part has convinced them that it is better to get partial results on 2 molecules in 24 hours than complete information on one molecule in 24 hours.

My opinion is that if we had enough computers to complete their project in a year or so, they would probably remove the time limit.

Lawrence

Thanks for the update Lawrence. What I am looking for is not necessarily removing the 12 hour limit for all computers processing the work for CEP2, but to allow individuals to set the maximum time they would want a WU to run. I have 4 computers (12 cores) running continuously since April 2010, and I would like to be able to provide all of the results for the 16 jobs in each WU I process instead of having the work skipped after the eleventh job.
----------------------------------------
[Edit 1 times, last edit by Former Member at May 30, 2013 4:53:22 PM]
[May 30, 2013 4:51:06 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Yarensc
Advanced Cruncher
USA
Joined: Sep 24, 2011
Post Count: 134
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Project always loses data

When it skips the ones after job 12 it isn't a time issue, its the program looking at the previous jobs and determining that there is nothing useful to be gained from continuing. Only the people over at Harvard would know what 'useful' would be. I think its safe to assume that they have it setup so that whatever your crunching at any given time is for the most beneficial to the advancement of the project.

yaren
[May 30, 2013 7:53:13 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 17   Pages: 2   [ Previous Page | 1 2 ]
[ Jump to Last Post ]
Post new Thread