Hello All,

I'm back in the fold after a few years away from crunching Mersennes.  I 
finally got a stable enough overclock with sufficient cooling to return 
numbers once again without error.

I've been assigned 4 exponents in the 41 million range (2560K FFT size), 
and P-1 factoring is underway.  It thinks the probability of finding a 
factor for these exponents is about 5.8% each.  By my calculations, 
though, P-1stage 2 factoring alone at its current pace will take 5.5 
days.   Stage 1 took just over 24 hours, for a combined total of 6.5 
days of factoring.   Benchmarks suggest I could run a complete LL test 
in 24 days.

Have I been away so long that it seems foolish now to spend 20% of total 
computing power on factoring when the odds of finding a factor are a 
small fraction of that?  Back when factoring was a few-hours-long 
process, it seemed like a good gamble to try, but at 6+ days for a 5.8% 
chance at saving 18 days' time it no longer seems like it.

QX6700 Quad Core 2.66 Ghz OC to 3.467 (passed 24 hours of torture testing)
Liquid cooling, CPU @ 60C under 100% load
4 GB Corsair Domaintor XMS PC2-8500C5D @ 1000 Mhz @ 2.2V
No overclock on the FSB -- 1066 Mhz (OCing the FSB corrupts the RAID)
Vista Ultimate x64
Prime-64 (x64 bit version) 24.14

I have indeed properly set core affinity for each instance of P64, and 
have allocated 128MB of RAM to each instance during daytime hours and 
384MB RAM to each instance overnight. 

After I wrote the above, a test indicated that if I raise RAM 
allocations to 384 MB around the clock, the estimated time to complete 
Stage 2 drops from 5.5 days to 1.5 days, so I've answered my own 
question for my own purposes... but given this, aren't the docs outdated 
when it comes to RAM requirements for P-1 factoring and/or the defaults 
for RAM allocation in the software such that most people won't be making 
effective use of their P-1 factoring time unless they both have the RAM 
to spare (as I do), and are educated enough to change the defaults?   
Are we at that point where just jumping to an LL test unless sufficient 
RAM is available for timely P-1 factoring might be prudent?  After all, 
I'd guess that many if not most participants aren't altering the RAM 
settings at all....

Jeff Woods






_______________________________________________
Prime mailing list
[email protected]
http://hogranch.com/mailman/listinfo/prime

Reply via email to