Hi, I understand the 321 sieve does not have very long left to run now before being ended. Does anyone happen to know roughly how much longer it will run for? I'm just curious - would also be nice to get the silver badge if there's time ;-)

Thanks - a useful post!

would also be nice to get the silver badge if there's time ;-)

4-6 Weeks!

This message explains the sieve is ending in 4-6 weeks. Get your badges NOW!

I'll stay with my gold, since it'd be hard pressed to get the next.


____________
My lucky numbers are 121*2^4553899-1 and 3756801695685*2^666669±1
My movie https://vimeo.com/manage/videos/502242

I think I can see why this is coming to an end (suspension).

I've been crunching for 2 days now, gone through 64 work units, and found 2 measly factors.

Using the "Recent average CPU time" as a benchmark (since I haven't done any 321 llr lately), it says a 321 sieve task takes 11:59:21, so my 64 WU would have taken 2,762,304 seconds. With a time of 17:16:02 per LLR work unit, in 2.7 million seconds I could have crunched 44 LLR work units.

So where LLR is eliminating about 22 candidates (when accounting for double checks), sieving is eliminating 2 candidates. Even if I am being extremely unlucky, and should have found TWENTY factors so far, we appear to be past optimal.

If anybody else has been keeping track of recent numbers of factors found, and has 321 LLR and sieve times, we can see how far off my estimates are.

CW

Ok. Update to my last post.

Since I started crunching last week, I've done 164 WU, and found 4 factors, for an average of 1 factor every 41 WU. The quickest sieve WU is taking about 7,738 (with the average one being somewhere around 7,775). 41 WU at 7738 seconds each is 317,258 seconds. I ran a 321 LLR on the same computer and it took 88,436 seconds.

This means in the time it took me to crunch the 41 WU required to get a factor, I could have processed 3.587 LLR work units. Since there are 2 WU necessary to complete a LLR, and sometimes more if there is a computation error, that 1 factor probably saves somewhere around 2.1 to 2.3 LLR tasks, so LLRing is still more time effective than sieving right now, but not nearly as much so as my previous post indicated.

Also, this is somewhat of a worst case. The computer I'm crunching both of these with is a Athlon 64 X2 6000+ running a Wubi install of 64 bit Linux, so it's getting the 64-bit benefit on the sieve WU, and the processor is probably on the poorer end of the spectrum when it comes to doing LLR as well.

So it still looks like a good decision to suspend the sieve, but not as much as it did the looking at the average CPU times.

The gods of PrimeGrid have spoken that the 321 Sieve barrel will never again be filled. No amount of prayer or graft can sway their decision.

It would be nice if everyone was as magnanimous as Pooh Bear 27, tocx and Steve* who have both stopped their 321 Sieving knowing they could not possibly reach the next badge.

Continuing to crunch drains the last remaining pool of candidates that we badge seekers desperately need.

I noticed one chap who already has a Ruby badge crunching away at a rate higher than any other member - please cease and desist so that we mere mortals can get that elusive silver or gold before the barrel is empty.

Nuff said.

Scott - Blunt is blunt - never a bit blunt :(

"The gods of PrimeGrid have spoken that the 321 Sieve barrel will never again be filled" is a humorous way of saying what Rytis was saying in his Post.

I do not question why other people are still crunching in the dying days of the 321 Sieve. I'm not sure that badges is the full why of my involvement - gosh, even anarchists do what they do for a reason. I simply asked consideration be given for badge chasers so that in the last few hours they have every opportunity for that last badge. No different from PrimeGrid asking people to change their focus towards some challenge goal each month.

So, to be more generic, on the off chance anyone reading this post really doesn't have a 321 Sieving target, please consider changing your preferences away from 321 Sieving so that people who want something (badges, credit targets etc) can get it. There are less that 13000 WUs left as at the time of my post.

I have my Silver now and I hope that other goal seekers get what they are actively seeking before the end.

Scott - Blunt is blunt - never a bit blunt :(

"The gods of PrimeGrid have spoken that the 321 Sieve barrel will never again be filled" is a humorous way of saying what Rytis was saying in his Post.

I do not question why other people are still crunching in the dying days of the 321 Sieve. I'm not sure that badges is the full why of my involvement - gosh, even anarchists do what they do for a reason. I simply asked consideration be given for badge chasers so that in the last few hours they have every opportunity for that last badge. No different from PrimeGrid asking people to change their focus towards some challenge goal each month.

So, to be more generic, on the off chance anyone reading this post really doesn't have a 321 Sieving target, please consider changing your preferences away from 321 Sieving so that people who want something (badges, credit targets etc) can get it. There are less that 13000 WUs left as at the time of my post.

I have my Silver now and I hope that other goal seekers get what they are actively seeking before the end.

Let me see - I have watched three previous projects come to the end of their life and can't remember PrimeGrid leaving it this late to top up the barrel. I have been watching the 321 Sieve sub-project status for the last few days and the rate of decrease suggests Rytis has planned very well that it will just hit zero on his target date. Nothing in the history or immediate past that suggests I don't get it.

In the absence of a last minute top up, having people grab work just to abort it is a bad idea. Assuming that they crunch some of it and no new work comes it will in fact make it less likely goal seekers will achieve their goals even if the get around to aborting what they don't process. Given Easter for many participants aborting might be delayed in the break.

Hopefully Rytis will chime in on this thread - proving you and Scott right and me wrong which would help everyone because of the possible last minute WU injection but until then I hold my view that if your not after a target, let those that are get on with their efforts.

Rytis - where are you? And bring an emergency barrel of fun!

When PSP sieve ended late last year they announced it would end on October 13th (I think), and they refilled the buffer for the last time on the 13th. People were still getting work for a solid 2 weeks after the last refill.

I guess I'm just assuming that the language they're using here means the same thing will happen here.

CW

Thank you Rytis for proving me wrong ;-)