PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
1) Message boards : Generalized Fermat Prime Search : PPS sieving, GFN 21 and 22 are all failing and I do not know why (Message 93623)
Posted 1640 days ago by Profile Steve Hawker*Project donor
All my GFN"s are failing and PPS sieving. I do not understand why. I have 3 Nvidia 680's set up thus PPS (CUDA), GFN 21 (cuda) and GFN 22 (open GL). I know I am fowling something up , but what. Any idea or directions are desired and would be greatly appreciated. Thanks


My GFN 18 (ati) tasks all now fail with

process exited with code 1 (0x1, -255)

This is after 172 straight valid tasks, approx 200 hours computing. I've changed nothing on my system.

Milky Way tasks run fine as they always did.

I notice a lot of PG apps were updated the other day, not atiGFN18 on Linux AMD though.
______________________________

I just tried atiGFN17MEGA tasks and got exactly the same thing. In the log I see:

Tue 29 Mar 2016 06:38:45 PM PDT | PrimeGrid | Started download of primegrid_genefer_3_3_0_3.12_x86_64-pc-linux-gnu__atiGFN17MEGA

Is this version 3.12? PG apps page says the latest should be 3.09 from last October.

If we've got a new app, looks like a library problem. Can someone confirm?

It seems we are having problems with the new Linux app. So it was reverted back to the previous version. You result might still have been from falty app, so I'd suggest trying another WU now and seeing if it's still failing.

Thanks for a speedy response.

Tried GFN17mega and GFN18 - both used 3.12 and both failed. I reset before trying.
2) Message boards : Generalized Fermat Prime Search : PPS sieving, GFN 21 and 22 are all failing and I do not know why (Message 93615)
Posted 1640 days ago by Profile Steve Hawker*Project donor
All my GFN"s are failing and PPS sieving. I do not understand why. I have 3 Nvidia 680's set up thus PPS (CUDA), GFN 21 (cuda) and GFN 22 (open GL). I know I am fowling something up , but what. Any idea or directions are desired and would be greatly appreciated. Thanks


My GFN 18 (ati) tasks all now fail with

process exited with code 1 (0x1, -255)

This is after 172 straight valid tasks, approx 200 hours computing. I've changed nothing on my system.

Milky Way tasks run fine as they always did.

I notice a lot of PG apps were updated the other day, not atiGFN18 on Linux AMD though.
______________________________

I just tried atiGFN17MEGA tasks and got exactly the same thing. In the log I see:

Tue 29 Mar 2016 06:38:45 PM PDT | PrimeGrid | Started download of primegrid_genefer_3_3_0_3.12_x86_64-pc-linux-gnu__atiGFN17MEGA

Is this version 3.12? PG apps page says the latest should be 3.09 from last October.

If we've got a new app, looks like a library problem. Can someone confirm?
3) Message boards : Number crunching : 2016 PrimeGrid Challenge Series ideas (Message 89751)
Posted 1776 days ago by Profile Steve Hawker*Project donor
Charley,

Thank you!!

I'm pleased to see that Cullen gets more than 5 days!

If you compare the durations to the current average run times, the subproject with the fewest WUs completed is Woodall, at just less than 2. If you extended the WOO/CUL challenge to 10 days, that should make managing resources more of a challenge as the average user could get 2 rounds in and maybe squeeze a few more bonus tasks in.

I would also suggest that the ESP-Sieve challenge is too long, if you want to shave off somewhere.

As for GFN, if you run just the smaller ones that might make it very interesting but I prefer the WR 15 day challenge we had in 2014.

Finally, would it be possible to update the standings immediately after the challenge with the warning that results are provisional until finalised. Waiting for 3 months for an update was so painful I had to work it out for myself and I'm a lazy dog.

Thanks again!!

s.
4) Message boards : Sicituradastra. message board (Message 86085)
Posted 1932 days ago by Profile Steve Hawker*Project donor
Actually the correct address is:

http://www.astrafamily.net

Could you delete the one you posted, please Sajjad?
5) Message boards : Sicituradastra. message board (Message 85742)
Posted 1952 days ago by Profile Steve Hawker*Project donor
Hi guys !

If you have been following the Enigma challenge closely, I guess you've noticed that SUSA has been ramping up HARD during the last 24 hrs.

We are kinda even with the output atm, but I think they'll push harder & still 3 days to go...

...so if ya have some spare cores ;)

(still 20 cores short & dunno when I'll get 'em fixed :( )

So, let's GO GUYS !!!



I just added 2 more cores

it all adds up!!
6) Message boards : Sicituradastra. message board (Message 85305)
Posted 1975 days ago by Profile Steve Hawker*Project donor
On the forum we had a thread for primes....

It's official! After three years of crunching, I've found my first prime number and it's a mere 992,038 digits long and ranks 134th largest prime number ever found. Wow!!

I found it at SRBase, not here at PG.

http://primes.utm.edu/primes/page.php?id=119824
7) Message boards : Sicituradastra. message board (Message 85248)
Posted 1978 days ago by Profile Steve Hawker*Project donor
FWIW, and I really don't know why I haven't tried this before, it really makes a difference if you reserve a CPU core for GPU tasks.

My 7950 was taking 90 hours or so for WR tasks but the one I've completed and the one I'm 25% through will come in at 76 hours each. Should be able to chuck in a couple of shorts. Sweet.

On the other hand, I cant get openCL tasks to run on my MacBook. OK, they run, but they soon crap out. Oh well.

Go Team!!
8) Message boards : Sicituradastra. message board (Message 85154)
Posted 1980 days ago by Profile Steve Hawker*Project donor
They don't give such good credits in challenge statistics. As I remember there was about 380000cr in past year, but because subproject has 50% long wu bonus we getting 600000 and even more in BOINC stats when task is validated.


I just had a look at last year's WR challenge stats. Seems like there were 841 tasks generating 489,160,910 cr for an average of 581,642.

I'd expect the numbers to be higher now because the k's are higher so tasks need more crunching.

In that challenge our top three were RaymondFO*, 288larsson and Ross* - I can't remember the last challenge Ray and Ross missed, if ever. So I have to conclude they are maxing out on the WR tasks.

As bro_david* said, all will be well....
9) Message boards : Sicituradastra. message board (Message 85140)
Posted 1980 days ago by Profile Steve Hawker*Project donor
I sure hope so Denis* :)

Have to run the short ones over here as the weather is very stormy atm and I just managed to get the boxes up before the challenge after 6 power outages during the evening, so I didn't want to risk loosing the big ones.

No outages so far after the start and the second batch will be done in a few hours, so I still have time to run some long ones on some boxes.

Does anybody know how much credits are granted for a long wu in the challenge ?


Wow! Fighting hard for the team!

Last time I ran a Wu I got just over 600,000 cr
10) Message boards : Sicituradastra. message board (Message 85122)
Posted 1981 days ago by Profile Steve Hawker*Project donor
Only 8 sicis on the board! A very low turn out comparing to the past, unless they are doing the BIG ones. I wonder if it has anything to do with the forum being down for so long......


i'll bet that everyone that can has chosen WR. It pays more.

We're well in the lead though so I wouldn't start worrying until day 4 when even my 7950 will have finished a WU. Then we'll get an idea of where all the firepower is. Hopefully ATP and CNT will still be out of the top 4.

Go Team!!


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2020 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.00, 0.01, 0.00
Generated 25 Sep 2020 | 17:02:11 UTC