20170512, 06:31  #23 
Romulan Interpreter
Jun 2011
Thailand
8932_{10} Posts 
I may also dedicate a Sandy to sieving assuming somebody comes out with a poly and a rangesplit (4 cores of i72600k).
I asked once, but no "expert" replied (and I am only guessing, as I never took such a job, neither dreamed to do something similar): how much memory would LA need? (I may have a box with 64G available for a while). Is actual LA multithreaded? How long time//core is expected for such a big LA job? 
20170512, 07:25  #24 
May 2009
Russia, Moscow
13×193 Posts 
I already did 4000@43e6 with no luck. I think we can run ECM up to t55 level and then ask yoyo for help with his army of minions

20170512, 11:34  #25 
Basketry That Evening!
"Bunslow the Bold"
Jun 2011
40<A<43 89<O<88
3×2,399 Posts 
I'll take a crack at 110e6 then (expect slow going)

20170512, 11:52  #26  
Jun 2012
2^{4}×181 Posts 
Quote:
So your 64 Gb machine should have enough memory, and presumably enough processing power etc to complete a 33bit job. Keep in mind this may be a 32bit job on the 16e siever. Either way, 64 Gb seems more than adequate. Check out the last few pages of this thread for some recent GNFS examples in the low 190s range. Last fiddled with by swellman on 20170512 at 11:53 

20170515, 12:42  #27 
Basketry That Evening!
"Bunslow the Bold"
Jun 2011
40<A<43 89<O<88
1C1D_{16} Posts 

20170522, 06:04  #28 
Basketry That Evening!
"Bunslow the Bold"
Jun 2011
40<A<43 89<O<88
3×2,399 Posts 

20170528, 19:13  #29 
Basketry That Evening!
"Bunslow the Bold"
Jun 2011
40<A<43 89<O<88
3·2,399 Posts 
Er, multiple typos in those posts. 500e6 was too much memory in S2 for my computer, so I reduced it to 250e6 while failing to report that here; meanwhile the 200 was a typo for 2000, so in sum I've done 4000 curves at 250e6 with no luck. I'm moving on to other projects for now. I guess this should go to yoyo?

20170531, 15:47  #30 
Oct 2006
Berlin, Germany
1001001111_{2} Posts 
How many curves at which B1 should I run?

20170531, 21:05  #31 
Jun 2012
2^{4}×181 Posts 
My best guess is 10,000 curves @B1=11e7 followed by 35,000 curves @B1=3e8.
Too much? Last fiddled with by swellman on 20170531 at 21:05 
20170601, 03:12  #32 
I moo ablest echo power!
May 2013
11011000111_{2} Posts 
Bayesian recommends
Code:
4700 curves with B1 = 850.0e6 2600 curves with B1 = 2.9e9 
20170601, 03:16  #33 
Oct 2006
Berlin, Germany
3×197 Posts 
No curves at 260e6, 850e6?
The 10,000 curves at B1=110e6 are now started. 
Thread Tools  
Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Reserved for MF  Sequence 4788  schickel  Aliquot Sequences  2911  20201124 18:01 
Reserved for MF  Sequence 3366  RichD  Aliquot Sequences  432  20201029 00:57 
Primes in nfibonacci sequence and nstep fibonacci sequence  sweety439  And now for something completely different  17  20170613 03:49 
80M to 64 bits ... but not really reserved  petrw1  Lone Mersenne Hunters  82  20100111 01:57 
A new sequence  devarajkandadai  Math  1  20070825 15:23 