Tidying up GIMPS
 

I'm not good at this.

But when I get an exponent I LL test it to completion.
I will P-1 beforehand (if necessary) and ask a friend to TF
a few extra bits on their shit hot GPU should it be needed.

Why are nearly all LL tests (to date) factored inadequately?
Why are there 10s of 1000s of 50M exponents hanging fire waiting for TF????

David

[QUOTE=davieddy;300032]Why are nearly all LL tests (to date) factored inadequately?
Why are there 10s of 1000s of 50M exponents hanging fire waiting for TF????[/QUOTE]What is the time frame that you are talking about?
Show your work to back up your assertions.:ermm:

[QUOTE=Uncwilly;300033][QUOTE=davieddy;300032]I'm not good at this.

But when I get an exponent I LL test it to completion.
I will P-1 beforehand (if necessary) and ask a friend to TF
a few extra bits on their shit hot GPU should it be needed.

Why are nearly all LL tests (to date) factored inadequately?
Why are there 10s of 1000s of 50M exponents hanging fire waiting for TF????

David[/QUOTE]

What is the time frame that you are talking about?
Show your work to back up your assertions.:ermm:[/QUOTE]

I think he is just making his biweekly/monthly dig at GPU72.

The problem though is more complex than a simple explanation. You have to consider 1) number of LL's requested per day 2) number of LL TF's completed per day by GPU72, 3) expiring exponents that are overdue, 4) Primenet's assignment strategy, 5) people who complain that they cannot get the TF's they want cause they are being hoarded and to try and satisfy them hundreds of under TF'd assignment are not picked up, yet due to #4 are again handed out for LL due to the inadequate TF requests by the complainers.

The GPU72 project is a huge success, but we are "running at 102 MPH in from of a train running 100 mph" and with all the overdue exponents getting tossed back into the mix, it feels like we are going 75mph and losing ground.

Maybe we need to petition George to tweak the assignment rules a bit, but looking at the exponents awaiting TF atm, everything the spider has grabbed under 60M is at 71 bits, so it's getting there.

[QUOTE=Uncwilly;300033]What is the time frame that you are talking about?
Show your work to back up your assertions.:ermm:[/QUOTE]
Post-Euler (by definition).
Not my work, but what the primenet summary displays.

David

Airplane
 

[QUOTE=bcp19;300039]I think he is just making his biweekly/monthly dig at GPU72.
[/QUOTE]
[B]Don't call me "Shortly";[/B]

I'm a [url=http://www.youtube.com/watch?v=6UZhPKydKcY]60 MINUTE MAN[/URL]

[QUOTE=bcp19;300039]...but we are "running at 102 MPH in from of a train running 100 mph" and with all the overdue exponents getting tossed back into the mix, it feels like we are going 75mph and losing ground[/QUOTE]

According to the empirical data available at Mersenne.info, in the last 30 days 18,592 candidates between 40M and 60M have been TFed to 72 or above bits. Or 620 a day.

During the same period and range, 6,174 LL tests completed, or 206 per day. So, to use your analogy, if GIMPS is running at 100 mph, then we're running at 300 mph.

Besides, at our current 30 day average rate, [URL="http://www.gpu72.com/reports/estimated_completion/primenet/"]we'll have TFed [B][I][U]everything[/U][/I][/B] to the newly appropriate levels below 60M in less than 106 days[/URL] (assuming we get access to everything currently not LLed, which isn't going to happen, so we'll be effectively finished even sooner).

I wonder what davieddy will harp about then?

[QUOTE=chalsall;300042]I wonder what davieddy will harp about then?[/QUOTE]

Um...., the next bit level.

[QUOTE=chalsall;300042]According to the empirical data available at Mersenne.info, in the last 30 days 18,592 candidates between 40M and 60M have been TFed to 72 or above bits. Or 620 a day.

During the same period and range, 6,174 LL tests completed, or 206 per day. So, to use your analogy, if GIMPS is running at 100 mph, then we're running at 300 mph.

Besides, at our current 30 day average rate, [URL="http://www.gpu72.com/reports/estimated_completion/primenet/"]we'll have TFed [B][I][U]everything[/U][/I][/B] to the newly appropriate levels below 60M in less than 106 days[/URL] (assuming we get access to everything currently not LLed, which isn't going to happen, so we'll be effectively finished even sooner).

I wonder what davieddy will harp about then?[/QUOTE]
Sheesh, it sounded SOOOOO good, but then you just had to go and ruin it with facts.:rolleyes: :lol:

[QUOTE=Prime95;300054]Um...., the next bit level.[/QUOTE]
Love you George.
Suitable link to follow as soon as it comes to mind.

D
Edit:
[url=http://www.youtube.com/watch?v=Sz-2jckjeHo]A little bit you[/url]
(Move quick for my age)

[QUOTE=bcp19;300055]Sheesh, it sounded SOOOOO good, but then you just had to go and ruin it with facts.:rolleyes: :lol:[/QUOTE]
[URL="http://www.youtube.com/watch?v=yX6FsTIq6ls"]Tense and nervous. Can't relax. Face up to the Fax[/URL]

PS Tina Weymouth on bass guitar.

[QUOTE=chalsall;300042]... so we'll be effectively finished even sooner)[/QUOTE]

Though, I might add, the LL leading edge in now in the 60-61M range so until GPUto72 is AHEAD there is still a need to forge on.