Results:
I tested with alim = 268M, rlim = 400M on the -a side, 33-bit large primes. Yield was high enough that I felt no reason to test higher lim's.
I tried mfbr of 65, 66, and 95; and mfba of 65,93,94,95,96. On the winning combo of 65/95 (65/94 was very close, likely within the margin of timing error), I tested lambdas of 2.7 and 3.7 vs 2.8 and 3.8; 2.7 and 3.7 produced the same yield on two different Q ranges, so I kept the smaller values.
Data, tested on regions of 1k:
[code]65/95, lambda 2.7/3.7:
Q=70M 0.557 sec/rel, 2523 rels (51 spQ)
Q=120M 0.633 sec/rel, 3208 rels (66 spQ)
Q=170M 0.754 sec/rel, 1666 rels (34 spQ)
Q=220M 0.713 sec/rel, 2248 rels (48 spQ)
Q=270M 0.821 sec/rel, 2604 rels (58 spQ)
Q=320M 0.809 sec/rel, 2517 rels (61 spQ)
Q=370M 0.903 sec/rel, 1772 rels (45 spQ)
Q=420M 0.836 sec/rel, 1512 rels (38 spQ)[/code]
I suggest sieving Q from 50M to 400M, expecting 750-800M raw relations.
I'm not certain 750M rels is enough for a job this size, but in gatekeepers I trust.

Here's the poly file and parameters used:
[code]n: 270190311360611008737597207914785628626394817649241238889983182977892667778914595601256771211961749590172206975699662088722691166582226058852847451219554336976835348606797148840109621091833378801
Y0: -46886185235594194332674628035591832985
Y1: 2261273527557887062
c0: 21723798235313825239630898066777328683391747732
c1: 2115241358681729821859927675022989718073
c2: 13959248941241170772739699872412
c3: -1257377794814468498747909
c4: -1186955160520804
c5: 4769856
skew: 126970960.11264
# lognorm 61.86, E 53.81, alpha -8.05 (proj -1.82), 5 real roots
# MurphyE = 1.02050678e-14
rlim: 400000000
alim: 268000000
lpbr: 33
lpba: 33
mfbr: 65
mfba: 95
rlambda: 2.7
alambda: 3.7[/code]

Nice work! Very thorough investigation. And a shoutout to Max for all his work as well. I think this is a new record poly for a C195, is it not?

Are you going to post or PM Tom, or do you want me to do it? Not sure it will be seen here.

Or is there any interest in a group sieve? It was discussed a while back but this job is a feasible 15e candidate. Either way I’ll help however I can.

Queued on 15e for NFS@home

C195 new best poly
 

@swellman, @VBCurtis
> I think this is a new record poly for a C195, is it not?
Posted it in the "Best msieve poly scores" ([URL]http://mersenneforum.org/showpost.php?p=481656&postcount=101[/URL]).

Is anyone working this sequence? It’s at a C146. And no I’m not volunteering - too many irons in the fire. :max:

Anyone?

If nobody posts about this composite in the next hour, I'll give it a T40 this evening and factor it overnight.

The C146 is cracked and posted, now on C196. I've done about half a T40, running curves at 20e6 presently. I'll take care of a t45, anyone interested should run 43e6 or bigger curves.

I'm nearly finished with a t50, where I will stop. Perhaps yoyo to finish ECM, the usual poly select folks, and 15e queue?

Yoyo is currently running t55, to be followed by a full t60 if necessary.

What is current status? I think this sequence has more chances to escape from driver than 3408 or 276.