RIP DCTF.

[Mark Rose]

Originally Posted by petrw1

May 10, 2016 ... we're going backwards.

I did miss 2.5 THz-days due to a power outage at the office last weekend. That's 4% of last week's work done, but that's only a third of the fall behind.

[petrw1]

Originally Posted by Mark Rose

I did miss 2.5 THz-days due to a power outage at the office last weekend. That's 4% of last week's work done, but that's only a third of the fall behind.

I suspect several contributors are falling into the "LetGPU72Decide" group and getting switched to LLTF.

On the bright side we do have at least 1 helper not in GPU72. tigerroars is taking 40M DCTF exponents to 72 bits.

[TheMawn]

May 2016 is a fairly long time from now. I had no idea it was going to take this long to finish (to be honest I was expecting more like May 2015 with a concentrated effort). I'd like to stay committed to RIP DCTF but I'm sure this is going to put some strain on LLTF.

If any LL's start getting released at 73 I'll start weaving in some LLTF.

[petrw1]

Agreed....while I will do what I can to finish DCTF if LLTF gets needy I'm there.

[tigreroars]

Originally Posted by petrw1

I suspect several contributors are falling into the "LetGPU72Decide" group and getting switched to LLTF.

On the bright side we do have at least 1 helper not in GPU72. tigerroars is taking 40M DCTF exponents to 72 bits.

shh. I'm supposed to be anonymous.

[petrw1]

Originally Posted by tigreroars

shh. I'm supposed to be anonymous.

Not if you are a real roaring tiger hehehe

[petrw1]

64,494 GhzDays DCTF in the last week. (About 4,000 more than last week!!!!)
63 different contributors
120 Factors found
7,213 P1/LL/DC work saved

22 contributors currently have assignments
3,587 Assignments out. (About 600 less than last week)

591.8399 estimated days to completion (About 3 months MORE than last week's estimate)
August 1, 2016 ... we're losing contributors :(

[chalsall]

Originally Posted by petrw1

591.8399 estimated days to completion (About 3 months MORE than last week's estimate)
August 1, 2016 ... we're losing contributors :(

That's not actually a bad thing (if the Project's goal is finding the next MP) IMHO.

We are currently about 81 days ahead of the DC "wavefront", and TF'ing about 100 DCs to 72 bits a day. As in, we're doing fine. Further, it should be noted that doing the "breadth first" search many are doing may be fun, but it doesn't really make all the much sense.

To be optimal, the DCTF'ers should be going to 72 bits in the 40M range. But LLTF'ing is really where the firepower is the most needed.

But, as always, people are free to do what they want....

[petrw1]

Originally Posted by chalsall

That's not actually a bad thing (if the Project's goal is finding the next MP) IMHO.

We are currently about 81 days ahead of the DC "wavefront", and TF'ing about 100 DCs to 72 bits a day. As in, we're doing fine. Further, it should be noted that doing the "breadth first" search many are doing may be fun, but it doesn't really make all the much sense.

To be optimal, the DCTF'ers should be going to 72 bits in the 40M range. But LLTF'ing is really where the firepower is the most needed.

But, as always, people are free to do what they want....

Is LL getting close to not keeping up?

Just to be clear on "breadth first" doesn't make sense. Are you implying it is better to:
1. Take all DC to 69, then 70 then 71 then 72
2. Take each lowest DC all the way to 72 ... this is where I currently am
3. Stop wasting time on DC and do LL instead

[Xyzzy]

What is the optimal LLTF strategy?

[Uncwilly]

Originally Posted by petrw1

Just to be clear on "breadth first" doesn't make sense. Are you implying it is better to:
1. Take all DC to 69, then 70 then 71 then 72
2. Take each lowest DC all the way to 72 ... this is where I currently am

1 is breadth (across the range)
2 is depth (taking each exponent to the goal depth).

I believe what Chris means is doing breadth might get you more factors sooner and you handle more exponents = more fun. While fully prepping exponents in order of need (lower first) is better for the long term. It makes things smoother.