Discussion related to cado-nfs

[paul zimmermann <Paul.Zimmermann@inria.fr>]

       Doug,

> Date: Tue, 16 Dec 2014 13:53:09 -0400
> From: "Douglas B. Staple" <dstaple@dal.ca>
> 
> Hello:
> 
> On the CADO-NFS website, I noticed a benchmark for CADO-NFS 2.1 factoring
> RSA-120 as 43.9 hours on a dual 8-core Intel(R) Xeon(R) CPU E5-2650.  This
> measurement is much slower than what I'm seeing on a less powerful system.
> Using a 2x Xeon E5540 (2 sockets, 4 cores per socket, 2 threads per core)
> running CentOS 6 with GCC 4.6.1 and GMP 5.1.0, it takes me about 6 hours to
> factor RSA-120 with Cado 2.1.1.  I am running a command of the form:
> ./factor.sh 
> 227010481295437363334259960947493668895875336466084780038173258247009162675779735389791151574049166747880487470296548479
> -t 2 -s 8
> 
> Is there a mistake on the CADO website, or is there something else strange
> going on?
> 
> Best,
> 
> Doug Staple

the time of 43.9 hours mentioned on the CADO-NFS website is the total
***cpu*** time, as explicitly said. This is the first figure you get
at the end of the factor.sh command:

Info:Complete Factorization: Total cpu/elapsed time for entire factorization: 
103.06/142.546

I guess the time of 6 hours you mention is the ***elapsed*** time.
If not, Houston, we have a problem...

Paul