20200603, 19:53  #67 
Nov 2003
1D24_{16} Posts 

20200603, 21:54  #68 
Jun 2015
Vallejo, CA/.
3BF_{16} Posts 
I agree.

20200604, 01:39  #69  
"Curtis"
Feb 2005
Riverside, CA
4243_{10} Posts 
Quote:
Or did you mean you wanted everyone else to do the work, while you watch for entertainment? 

20200604, 15:06  #70  
Nov 2003
2^{2}×5×373 Posts 
Quote:
previously making himself heard. 

20200606, 11:12  #71 
(loop (#_fork))
Feb 2006
Cambridge, England
6353_{10} Posts 
I'm unsure as to whether R323 is accessible with the 16e siever. It seems that x^610 sieves slightly better than the octic, but with a yield of about 0.8 relations per Q the search range is annoyingly large. Will compare 33/34/35bit large primes for the sextic under CADO; for the octic the yield boost going to 35bit LP isn't enough.
Last fiddled with by fivemack on 20200606 at 11:14 
20200606, 12:57  #72  
Nov 2003
7460_{10} Posts 
Quote:
The last is 4 digits larger. I expect that numbers up to ~330 digits should be doable by NFS@Home. Perhaps you are not allowing the 16f siever? Do you really mean to restrict which siever is used? 

20200606, 15:53  #73 
Sep 2009
747_{16} Posts 
Would the 16f siever support j 16? That boosts yield on 16e (It makes the sieve area 16k x 16k) but raises memory required.
Chris 
20200607, 03:06  #74 
"Curtis"
Feb 2005
Riverside, CA
4,243 Posts 
When I tested J 16 a couple years ago, it worked on some numbers and Qranges but not others. I didn't find a pattern in which polys were resilient, and which crashed a bunch, so I deemed it unusable.
That 40% yield boost would clearly extend GGNFS sievers another 68 digits! 
Thread Tools  
Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
7+ table  garo  Cunningham Tables  85  20200415 21:12 
5 table  garo  Cunningham Tables  82  20200315 21:47 
5+ table  garo  Cunningham Tables  99  20200110 06:29 
6+ table  garo  Cunningham Tables  79  20200101 15:26 
6 table  garo  Cunningham Tables  41  20160804 04:24 