[QUOTE=pinhodecarlos;524838]Q4. Yes, it’s lasieve5f application, 16e Lattice Sieve V5 on the project status. I can attach the windows binary used there not the Linux one.[/QUOTE]

5I16e not 5I16f I believe. nfs@home uses a confusing naming system. 5I16f can sieve composite special q

I use 16e, not 16f. There was an issue with 16f when I deployed the sievers and never changed them. The GNFS parameters for 5th order polynomials are

lss: 0
alim: 250000000
rlim: 250000000
lpbr: 33
lpba: 33
mfbr: 67
mfba: 96
rlambda: 2.6
alambda: 3.6

Swap a and r parameters for sieving on the rational side. These are chosen to provide decent yield with relatively low memory use.

[QUOTE=swellman;524815]
User e-score post # deg spin?

VBCurtis 1.398E-15 47 6 yes

vebis 1.188E-15 23 5 yes
Gimarel 1.181E-15 61 5 yes
Gimarel 1.176E-15 42 5 yes
[/QUOTE]

I've only tested these so far, but each deg 5 is at least 20% faster than the top-scoring deg 6. Each of the deg 5 have something going for them- the fastest is the lowest-yielding, and vice versa. Once I test a couple Q ranges from each poly to narrow down to a top 3-4, I'll sample 8 or 10 different ranges to see if the slower speed is a good tradeoff for a smaller Q-range (the idea being that avoiding the highest Q's means also avoiding the slowest sieve ranges).
Using Greg's params (except I replaced 67 with 66 for LPB) I'm getting timings around 1.2 sec/rel and yields of 1.2 to 1.5. I've only tested Q=100M and 500M so far, though; much more to come.

NFS@Home almost done with sieving, last day my machine has been handling left overs from other clients not processing wus on time.

It split into P75 (just like 2,2330L) x P136.