Got first batch of TF to 75

[Chuck]

I received my first batch of exponents via MISFIT to TF to 75 bits (66M). Knew this was coming (let GPU72 decide)...I have to get used to a four-hour run time now.

[chalsall]

Originally Posted by Chuck

I received my first batch of exponents via MISFIT to TF to 75 bits (66M). Knew this was coming (let GPU72 decide)...I have to get used to a four-hour run time now.

Yes. Based on our current trending, we should /just/ be able to pull this off...

Those who have recently stopped or lessened their TF'ing altogether, or are currently DCTF'ing, it would be greatly appreciated if you'd help build a bit of a buffer for LL Cat 3 (where most of the LL'ing is currently being done; ~150 candidates a day).

Even only going to 74 in the Cat 4 range would be useful; that's where the P-1'ers are currently working.

So everyone knows, we're currently about 2.5 months ahead in the DC range (to optimal levels), and are about to complete LLTF levels optionally for Cat 2.

(And to think a certain David was certain we couldn't sustain TF'ing to 73.... )

[Chuck]

Originally Posted by Chuck

I received my first batch of exponents via MISFIT to TF to 75 bits (66M). Knew this was coming (let GPU72 decide)...I have to get used to a four-hour run time now.

My completed TFs to 75 are being picked up for LL testing just about as fast as I finish them.

[petrw1]

Originally Posted by Chuck

My completed TFs to 75 are being picked up for LL testing just about as fast as I finish them.

Mine too....and not surprisingly CC got most of then

Think it means Spidey is getting smarter :)

[Chuck]

Originally Posted by Chuck

I received my first batch of exponents via MISFIT to TF to 75 bits (66M). Knew this was coming (let GPU72 decide)...I have to get used to a four-hour run time now.

The work to 75 bits doesn't seem to be showing up on the Workers' Overall Progess page, although the totals continue to accumulate. Especially noticeable on the progress for the last day, where there is a total but no counts in any of the other columns.

[Xyzzy]

We have declined to take on TF75 work until we can be certain that the assignments we get are exclusively ours.

In this week's batch of TF74 work we lost 4 GPU hours. If those had been TF75 assignments we would have lost 16 GPU hours.

[Chuck]

Originally Posted by Xyzzy

We have declined to take on TF75 work until we can be certain that the assignments we get are exclusively ours.

In this week's batch of TF74 work we lost 4 GPU hours. If those had been TF75 assignments we would have lost 16 GPU hours.

These were completed by user A_Savchenko; is there something fishy going on with his assignment procedure?

I haven't seen any "not needed" results from my TF to 75 efforts yet.

[chalsall]

Originally Posted by Chuck

The work to 75 bits doesn't seem to be showing up on the Workers' Overall Progess page, although the totals continue to accumulate. Especially noticeable on the progress for the last day, where there is a total but no counts in any of the other columns.

Thanks for pointing that out.

Assigning work to 75 was easy; there are still some reports which need to be adjusted to reflect this, such as the Workers Progress and Cost reports.

Edit: Please do know that all work to 75 (or beyond) *IS* being recorded, as can be seen on your individual completion reports as well as the Overall System Progress graphs. I just need to expand the Worker's batch query and associated report, and some other reports.

[chalsall]

Originally Posted by Chuck

These were completed by user A_Savchenko; is there something fishy going on with his assignment procedure?

Hmmm... It appears that GPU72 expired this TF'ing assignment to A_Savchenko after 45 days, which Xyzzy then picked up. A_Savchenko then completed the TF assignments.

Sorry about that. I'm going to fully disable the recycling script for the time being (even with the extended period allowed) and will try to contact those who have outdated assignments before recycling (as my limited time allows).

Please rest assured, however, that GPU72 does NOT assign candidates for TF'ing to more than one person at a time (unless recycled, or released by the worker).

[Chuck]

M66250619 has a factor: 32217303685264626845711 [TF:74:75*:mfaktc 0.20 barrett76_mul32_gs]

Finally I found one; it has been nearly two weeks without a positive result. What with Oliver doing the P-1, the decreased probability and longer run times, this is probably going to be what to expect going forward.

[TheMawn]

Originally Posted by Chuck

M66250619 has a factor: 32217303685264626845711 [TF:74:75*:mfaktc 0.20 barrett76_mul32_gs]

Finally I found one; it has been nearly two weeks without a positive result. What with Oliver doing the P-1, the decreased probability and longer run times, this is probably going to be what to expect going forward.

I've got my stuff working on 75 bits too and it's soooo sloooowwwww. Maybe it's bad luck but maybe not. I can only clear about 6 per day per card so to hit 75 attempts would take almost two weeks.