GPU to 72 status...

[Aramis Wyler]

Originally Posted by chalsall

... many more assignments than normal are being handed out by Primenet since the "big news".

I trimmed my lltf from 75 to 74 bits for the sake of speed just in case we get overtaken in the 59M block.

[kladner]

Originally Posted by bcp19

12 factors in one day :) 8 - 33M and 4 - 61M out of 300 tested exponents. I think I hit the jackpot!

"I don't know. It must have been the roses." -GD (or the moon, or Mercury, or something.....)

On Feb. 6 I pulled down 3 factors out of 43 tested. All mid-61M, 25 71-73, 18 72- 73 (one of the factors in the latter group.)

I had to study a bit to get the numbers right. By GPU72's clock the factors landed at ~00:45 on the 7th, which, of course is still in progress here at -6 UTC. However, by my local time that is 18:45. My other LLTF upload on the 6th was at ~14:45 UTC, or 08:45 CST (US).

Figuring by percentages then

Jerry = 2.667%
Kieren = 6.977% ()
Both = 3.207% factors

I wonder if anyone else hit the jackpot on the 6th, or what the factor percentage for GPU72 was compared to the long-term average.

[chalsall]

Originally Posted by kladner

]I wonder if anyone else hit the jackpot on the 6th, or what the factor percentage for GPU72 was compared to the long-term average.

Yesterday was a very good, but not an exceptional, day. Please see this graph.

Note that because of GPU72's temporal spreading algorithm, most of the factors reported on the 6th are shown as the 5th here. As in, the candidates for which factors were found were assigned on the 5th or earlier.

The Linear Regression Trend line is the most important to follow.

[kladner]

Thanks! I see what you mean. The upward trend is good to see. Also, 1/31 was even better, especially in P-1.

[petrw1]

If I look at this report:

http://www.gpu72.com/reports/estimat...tion/primenet/

It suggests that by the end of 2013 all the required DC-TF to 50M and (by my extrapolation) all the LL-TF to 70M as well will be complete

Then what?

I'm sure neither wave will be anywhere those limits by then.

[kladner]

Originally Posted by petrw1

If I look at this report:

http://www.gpu72.com/reports/estimat...tion/primenet/

It suggests that by the end of 2013 all the required DC-TF to 50M and (by my extrapolation) all the LL-TF to 70M as well will be complete

Then what?

I'm sure neither wave will be anywhere those limits by then.

Redirect to LL testing? LMH factoring? Will P-1 also be caught up at that point?

[ckdo]

Originally Posted by petrw1

by the end of 2013 all the required DC-TF to 50M and (by my extrapolation) all the LL-TF to 70M as well will be complete

Your link is broken. Anyway, the ETAs are of the "if everyone does only this" type.

Since DCTF only gets around 1/7th of GPU72's TF power, at least on that front we are save for a few years. We'll probably try and close the gap to the LLTF tail afterwards.

As far as LLTF goes, we'll probably simply go higher, or deeper, or both.

[c10ck3r]

Am I correct in assuming that once the DC wave is where the LLTF wave began, there will be no more DCTF?

[chalsall]

Originally Posted by c10ck3r

Am I correct in assuming that once the DC wave is where the LLTF wave began, there will be no more DCTF?

Mostly, yes.

[kracker]

Originally Posted by c10ck3r

Am I correct in assuming that once the DC wave is where the LLTF wave began, there will be no more DCTF?

I think there will always be a Double Check TF wave... it just goes up, unless they all get factored before first LL... which probably will happen.

[chalsall]

Originally Posted by kracker

I think there will always be a Double Check TF wave... it just goes up, unless they all get factored before first LL... which probably will happen.

No.

Because a DC candidate only needs about one "bit" less of TFing than a LL candidate in about three years or so there will be no DCTF left to do.