Discussion related to cado-nfs

["Thomas D. Dean" <tomdean@speakeasy.org>]

Does CPU time for sieve include the time for all threads?

I lost track of elapsed time.  The machine was lightly loaded other
than the factor code.  Is the elapsed time approximately the sum of
the time reported divided by the number of cores?

I used factor.sh with a 135 digit number and -t 10 on an Intel
i7-3930K at .2GHZ.

# factor.sh 
296603151077074197308684124715622560525294501766069587161483751777320391554230009758029564174782423278870115094641644418265512090475910
 -t 10
...
Info:--------------------------------------------------------------------------
Info:All done!
Info:--------------------------------------------------------------------------
2 5 2122943609450249057621 405620061201508643988375124643817406843652803 
34444340671702467488860342883734479061576134487870291451728421045657
OK
CPU time for polyselect:      19h31m19s
CPU time for sieve:           33d9h16m59s
Real time for linear algebra: 11h0m22s - 1h36m22s - 6h6m47s (krylov - lingen 
- mksol)
CPU time for sqrt (one root): 28s - 4m50s  (ratsqrt - algsqrt)

The machine was up almost 10 days.

# uptime
 9:50AM  up 9 days, 21:31, 4 users, load averages: 6.94, 5.14, 3.43

tomdean