Jason's sieved ranges(for the prp and llr addicts with Intels)

[jasong]

Okay, guys, I've just discovered the upload option.

I'm unreserving 74632 and supplying the sieve file. I'm going to start at the top of the list and sieve the numbers in order until it's 1 k every 35 minutes, then upload the result. I'm running a 2200+ Sempron, so if anyone thinks I should change my sieving rule, please post.

Reserved 2822 and 3706 (Sorry, wrong thread for reserving)

[geoff]

I will prp this file (k=74632), thanks.

The largest candidate in the file, 74632*5^199950+1, takes about 42 minutes (5.45ms/bit, 48K FFT, 2.66GHz P4 with PRP3).

[jasong]

This 35 minute rule is causing my Obsessive Compulsive Disorder to use and abuse me, so I'm changing the rule to,"Sieve to 100 billion," and make creative use of Priority settings so I don't feel the need to get up in the middle of the night to check the computer.

Mental illness sucks.

[geoff]

Sieving to 100 billion might be OK for the candidates with very heavy Nash weights, but is probably a waste for the lighter candidates. (assuming nmax=200,000).

The 'sieve until the rate is x seconds per n' allows for the weight differences automatically, as lighter candidates will dry up earlier. You can get NewPGen to automatically start prp testing with PFGW once it finishes sieving so that no machine time is wasted, or run another task at lower priority that will take over once NewPGen stops.

[jasong]

2822, 1 n every 40 min 88.3 billion
3706 1n every 35 min 51.2 billion
(I'm going to use Geoff's advice and set things to 40 min)

Thanks dude.

[jasong]

Only one file can be edited at once(Learn something new every day).

[jasong]

6082 sieved file(sieved to 46.7 billion)

[jasong]

sieved to 57.6 billion

[geoff]

I have put these sieve files in the Sierpinski5/sieve/ directory, this way they can be edited to remove tested values. If anyone else has sieve files they want to make available, send them to me by PM or email and I will link them from the reservations threads.