GPU to 72 status...

[ET_]

Originally Posted by kladner

Fascinating! Thanks for sharing.

Tidy!

Luigi

[chalsall]

Originally Posted by Xyzzy

Would you be willing to make the columns of data sortable

Rather than sortable columns, how about different reports for each work type?

Also, I've changed the query slightly for the Workers' Progress reports so that if someone works from, for example, 69 to 72 the intermediate steps are shown. As in, there's a +1 at 70, 71 and 72 rather than just at 72.

This makes much more sense, and helps make obvious why Craig is still ahead of Xyzzy even though Xyzzy has a slightly higher number of completions to 72.

[James Heinrich]

Originally Posted by chalsall

Rather than sortable columns, how about different reports for each work type?

Good, but introduces a bug you may not have thought of:
The P-1 report for example shows Chris Halsall at #22, with a link to the "Overall Work Done by Workers ranked 21 to 30" graph. Problem is, you're ranked #22 for P-1 work, but not #22 overall, which is what the linked graph shows.

[chalsall]

Originally Posted by James Heinrich

Good, but introduces a bug you may not have thought of:

Yeah, I had thought of that. And it will be fixed once I have work type specific graphs....

[chalsall]

Hey all.

As of right now, the GPU to 72 Project has found 600 factors.

208 P-1 tests saved.
1,082 LL/DC tests saved.
91,053 GHz Days (249 GHz YEARS) saved.

Thanks for all the cycles everyone!!!

[petrw1]

Originally Posted by chalsall

Hey all.

As of right now, the GPU to 72 Project has found 600 factors.

208 P-1 tests saved.
1,082 LL/DC tests saved.
91,053 GHz Days (249 GHz YEARS) saved.

Thanks for all the cycles everyone!!!

And may I add over 100 of those factors are from us P-one-ers.

[chalsall]

Originally Posted by petrw1

And may I add over 100 of those factors are from us P-one-ers.

Indeed!!! 104 factors found by way of P-1 efforts, out of 2,155 runs.

At first we were having to return candidates back to PrimeNet without P-1 work being completed because of lack of participation. Now the system can't even keep it's cache of 250 full!

[petrw1]

Originally Posted by chalsall

Indeed!!! 104 factors found by way of P-1 efforts, out of 2,155 runs.

At first we were having to return candidates back to PrimeNet without P-1 work being completed because of lack of participation. Now the system can't even keep it's cache of 250 full!

Gee, I hate to drop in a lump of coal in your stocking so close to Christmas but I have decided to switch work types in 2012 (New Year --> New Goal). I will be doing GPU-DC work starting in the new year on at least 13 cores. By then I will have completed 600 GPU P-1s....let's hope there will be others to take over my share of P-1 (15 per day).

[kladner]

Originally Posted by petrw1

..... I will be doing GPU-DC work starting in the new year on at least 13 cores. By then I will have completed 600 GPU P-1s....let's hope there will be others to take over my share of P-1 (15 per day).

Quibble: Surely those 600 P-1's are CPU? Just asking.

[petrw1]

Originally Posted by kladner

Quibble: Surely those 600 P-1's are CPU? Just asking.

Yes, I am GPU-less.

3 Quads and 1 PIV full time P-1 for almost 2 months

[James Heinrich]

Originally Posted by kladner

Quibble: Surely those 600 P-1's are CPU? Just asking.

I'm pretty sure by "GPU-P1" and "GPU-DC" petrw1 is referring to assignments from www.gpu72.com as opposed to actually doing the work on a GPU.

I'll pick up a small amount of the slack by adding 4 cores of a i7-3930K into the mix for P-1 (the other 2 cores are busy feeding the GTX570 for TF ).