The "(W)TF Depth Debate" state of the bunion

[Prime95]

Originally Posted by davieddy

Reason: rudeness.

Merry Christmas. Your post in the moderation queue starting a new thread about whether LL users should have access to the smallest exponents will not be approved. Your demand that you be taken off moderation is a non-starter. As to your recent insulting behavior, this is my notice to you that your next rude and insulting post will result in a lifetime ban. I've never done that before - you can be my first.

[chalsall]

Originally Posted by chalsall

A bit less than 11,000 LL candidates below 57M are only TFed to 72 (read: most of them). With the average "abandonment rate" of ~ 80%, we could process most of them over time. Although, I imagine this rate will be lower because of the "preferred" status of the very lowest range.

As many of you will have noticed, GPU72 has begun the process of bringing in recycled candidates between 54M and 57M TFed to 72 "bits", and making them available for TFing to 73.

I have also updated the Estimated Completion Report to reflect this new effort. At our current 30 day rolling LLTFing rate, we could process the 6,919 candidates in this range in 9.8 days (~700 candidates per day!). Of course, we won't get access to all of these, and those we do only over the next two months or so (a hundred or so a day).

Also, while I'm writing, we've almost finished off bringing 62M up to 74 from 73. An extra 165 candidates eliminated by doing this work. Thanks for the suggestion David!

Edit: While I was typing this Oliver dumped his week's work. This now fully completes taking 63M to 74!

[Mark Rose]

Originally Posted by chalsall

As many of you will have noticed, GPU72 has begun the process of bringing in recycled candidates between 54M and 57M TFed to 72 "bits", and making them available for TFing to 73.

I would like to chew on those, however the Get Assignments page won't let me! This might be related to the hack you put it to stop me from getting the 50M exponents a month or two ago.

[davieddy]

Originally Posted by chalsall

Also, while I'm writing, we've almost finished off bringing 62M up to 74 from 73. An extra 165 candidates eliminated by doing this work. Thanks for the suggestion David!

No problem.

I just thought that it was work which had to be done eventually, and would help meet the demand for LL assignments.

[davieddy]

Originally Posted by chalsall

Also, while I'm writing, we've almost finished off bringing 62M up to 74 from 73. An extra 165 candidates eliminated by doing this work. Thanks for the suggestion David!

No problem.

FB told me so many times that that TF to 74 >63M was getting so far ahead of LL assignment that I drew the logical conclusion that it must be true.
At first I thought "That's a Good Thing^TM. Let's keep up the good work until the lead is big enough to contemplate starting on 75 bits, even though initially it won't keep pace".
The I thought "No. George and Chris want to tidy up their legacy <63M. Perhaps I should encourage them. Then they will love me even more than they do already".

[kladner]

Be sure that you set all the fields appropriately. I had a problem until I remembered to reduce the "Factor To" value to 73. You might also try setting an upper range limit, and/or selecting "Lowest Exponent".

[chalsall]

Originally Posted by kladner

Be sure that you set all the fields appropriately. I had a problem until I remembered to reduce the "Factor To" value to 73. You might also try setting an upper range limit, and/or selecting "Lowest Exponent".

Correct. Set the "To" value to be 73, and select "Lowest Exponent", and the 50Ms will appear in the preview panel (if any are available, of course).

[Mark Rose]

Originally Posted by kladner

Be sure that you set all the fields appropriately. I had a problem until I remembered to reduce the "Factor To" value to 73. You might also try setting an upper range limit, and/or selecting "Lowest Exponent".

That did it :)

[davieddy]

Originally Posted by Prime95

This doesn't pass a simple common sense test. Pose the question this way: Would GIMPS be better off if your GPU took two 58M exponents from 72 to 73 or one 64M exponent from 73 to 74?

1) The 2 58M TFs take only a little more time than the 1 64M TF.
2) The 2 58M TFs have a little more than twice the chance of finding a factor.

I claim the extra iterations and higher FFT length of a saved 64M LL test does not compensate for the nearly double chance of saving one 58M LL test.

I would say the overriding decider was that taking the 64M expo to 74 bits created a virgin LL assignment.
Furthermore neither are worth the effort.

Originally Posted by Prime95

Merry Christmas. Your post in the moderation queue starting a new thread about whether LL users should have access to the smallest exponents will not be approved. Your demand that you be taken off moderation is a non-starter. As to your recent insulting behavior, this is my notice to you that your next rude and insulting post will result in a lifetime ban. I've never done that before - you can be my first.

Perhaps you and Chris might delegate your moderation to others - Phil, Paul and William spring to mind - since you obviously have a conflict of interests in this thread.

BTW a lifetime ban sounds exceptionally lenient in my case

D

[kladner]

Originally Posted by Mark Rose

That did it :)

Cool!

[davieddy]

Originally Posted by kladner

Cool!