It's been announced over at S@H, that ABC will shortly be shutting down. Cannot find any info on front page regarding this.
Is this true & if so, when?
____________
/ 
It's been announced over at S@H, that ABC will shortly be shutting down. Cannot find any info on front page regarding this.
Is this true & if so, when?
Click the Rekem mee met ABC link on this project's front page. Then, if you have the Flashplayer plugin, on the left side of the screen you'll see ABC's progress. Today it's reading just over 74% complete.
This is ABC's first run. In this run we are searching the number space 1 to 10^18 for ABC triples. It looks like we may be done in early 2009. Who knows, after that they may decide to do a second run and search a larger number space? Or they may decide to end the project. So far there has been no official word.
/ It's been announced over at S@H, that ABC will shortly be shutting down. Cannot find any info on front page regarding this.
Is this true & if so, when?
Click the Rekem mee met ABC link on this project's front page. Then, if you have the Flashplayer plugin, on the left side of the screen you'll see ABC's progress. Today it's reading just over 74% complete.
This is ABC's first run. In this run we are searching the number space 1 to 10^18 for ABC triples. It looks like we may be done in early 2009. Who knows, after that they may decide to do a second run and search a larger number space? Or they may decide to end the project. So far there has been no official word.
Thanks Dagorath.
____________
/ 
The project has been running since January 2007, so the 75% took almost two years. There might be an increase in crunching power (some people might decide to crunch ABC@home now, because it's getting closer to an end, and Santa Claus might bring some new computers ;)), but I don't think it will be finished before January or February 2009.
____________
Proud member of SETI.Germany
/ 
The project has been running since January 2007, so the 75% took almost two years. There might be an increase in crunching power (some people might decide to crunch ABC@home now, because it's getting closer to an end, and Santa Claus might bring some new computers ;)), but I don't think it will be finished before January or February 2009.
+2% per week?
/ The project has been running since January 2007, so the 75% took almost two years. There might be an increase in crunching power (some people might decide to crunch ABC@home now, because it's getting closer to an end, and Santa Claus might bring some new computers ;)), but I don't think it will be finished before January or February 2009.
+2% per week?
I have tried to keep a graph of the progress we have made but I haven't done a good job of it, I don't have many data points. The long straight part is evidence of many months of no data, not a very informative graph. However, the graph does show the rate of progress has accelerated significantly. It looks like the previous 25% required 5 months so +2% may be a little too optomistic. On the other hand, when ABC first went into production I recall Hendrik saying the work units would get shorter as we progress. Maybe finish this run near end of March 2009?
/ 
The project has been running since January 2007, so the 75% took almost two years. There might be an increase in crunching power (some people might decide to crunch ABC@home now, because it's getting closer to an end, and Santa Claus might bring some new computers ;)), but I don't think it will be finished before January or February 2009.
I can tell you some of us have come over with some of our horsepower because of other Projects having some issues. I for instance came to ABC in August of 2008 bringing most of my dual, or more, cores with me. I am guessing but that is probably more than left during that same time period.
____________
/ It's been announced over at S@H, that ABC will shortly be shutting down. Cannot find any info on front page regarding this.
Is this true & if so, when?
ABC is not shutting down. When we reach our inital goal we will increase the boundary to probably 10^21. This might take some time as we need a new application as numbers get larger than 2^64 and there's no C type for integers larger than 64 bit. So we are gonna write a new application which can handle integers > 2^64. Because the numbers are larger than 2^64 the application will be a little slower, but because of algorithmic inprovements, the application will be quite a bit faster in total.
____________
"Life is short and meaningless, unless you make the best of it."
/ That's good news :)
/ It's been announced over at S@H, that ABC will shortly be shutting down. Cannot find any info on front page regarding this.
Is this true & if so, when?
ABC is not shutting down. When we reach our inital goal we will increase the boundary to probably 10^21. This might take some time as we need a new application as numbers get larger than 2^64 and there's no C type for integers larger than 64 bit. So we are gonna write a new application which can handle integers > 2^64. Because the numbers are larger than 2^64 the application will be a little slower, but because of algorithmic inprovements, the application will be quite a bit faster in total.
Thank you very much, that is EXCELLENT NEWS!!
____________
/ It looks like the previous 25% required 5 months so +2% may be a little too optomistic. On the other hand, when ABC first went into production I recall Hendrik saying the work units would get shorter as we progress. Maybe finish this run near end of March 2009?
See my initial post from 29th of August: http://abcathome.com/forum_thread.php?id=80&nowrap=true#8952 where I predicted 1st of April 2009.
ABC is not shutting down. When we reach our inital goal we will increase the boundary to probably 10^21. This might take some time as we need a new application as numbers get larger than 2^64 and there's no C type for integers larger than 64 bit. So we are gonna write a new application which can handle integers > 2^64. Because the numbers are larger than 2^64 the application will be a little slower, but because of algorithmic inprovements, the application will be quite a bit faster in total.
Even when there will be a new target, there will be a break in production, because of programming and testing the new application. Perhaps a revival for the ABC@Home Beta ;)
/ ABC is not shutting down. When we reach our inital goal we will increase the boundary to probably 10^21. This might take some time as we need a new application as numbers get larger than 2^64 and there's no C type for integers larger than 64 bit. So we are gonna write a new application which can handle integers > 2^64. Because the numbers are larger than 2^64 the application will be a little slower, but because of algorithmic inprovements, the application will be quite a bit faster in total.
How long will the pause in the project be and will the current credits be transferred?
/ How long will the pause in the project be and will the current credits be transferred?
We'll try to make this transition as smooth as possible, but there might be a short outage, it's too early to tell now. Since we're running the same project with a different app and larger workunits, the credit will remain intact.
____________
"Life is short and meaningless, unless you make the best of it."
/ How long will the pause in the project be and will the current credits be transferred?
We'll try to make this transition as smooth as possible, but there might be a short outage, it's too early to tell now. Since we're running the same project with a different app and larger workunits, the credit will remain intact.
Thanks for the clarification and quick response!
/ Cleaned up the graph of ABC progress I've been keeping. You can actually read the dates on the x-axis now. Both the x and y axes are extended to allow extrapolation of the completion date. If the current rate of progress holds, this run will be done before 02/2009.
The URLs for the data table and gnuplot script used to create the graph are at the bottom of the graph. The first dozen or so point were gleaned from the totale project status thread where there are irregular reports of the progress on the rekenmee site. I'll try to collect the % progress from rekenmee every few days from now on.
/ Cleaned up the graph of ABC progress I've been keeping. You can actually read the dates on the x-axis now. Both the x and y axes are extended to allow extrapolation of the completion date. If the current rate of progress holds, this run will be done before 02/2009.
The URLs for the data table and gnuplot script used to create the graph are at the bottom of the graph. The first dozen or so point were gleaned from the totale project status thread where there are irregular reports of the progress on the rekenmee site. I'll try to collect the % progress from rekenmee every few days from now on.
Progress does seem to be accelerating lately.
____________
/ We'll try to make this transition as smooth as possible, but there might be a short outage, it's too early to tell now. Since we're running the same project with a different app and larger workunits, the credit will remain intact.
Hi Robin,
it seems we will reach 100% next month :-) How far are you with the development of new application? Can we expect a smooth transition to new phase or will you need more time for preparation?
Thanks for info & Happy New Year :-)
H99
/ ABC is not shutting down. When we reach our inital goal we will increase the boundary to probably 10^21. This might take some time as we need a new application as numbers get larger than 2^64 and there's no C type for integers larger than 64 bit. So we are gonna write a new application which can handle integers > 2^64. Because the numbers are larger than 2^64 the application will be a little slower, but because of algorithmic inprovements, the application will be quite a bit faster in total.
Would that mean anything for chrunching speed @ 32-bit vs 64-bit computers? 64-bit is twice at fast because those 64-bit integers fit in one register (or something like that, I forgot), but with larger integers would be slower too... So instead of 100% boost 64-bit would increase chrunching speed with 66%, or doesn't it work like that?
/ I've started a new thread about the continuation of the project.
-Willem Jan
/