[Coin-discuss] Cbc's great parallel performance

[Sebastian Nowozin]

Hello everybody,

there is a new MILP benchmark result at Hans Mittelmann's benchmark page,

    http://plato.asu.edu/ftp/milpf.html

He compares Cbc 2.3 with 1/2/4 CPUs and for most models this improves
the solution speed quite noticably.

    Good job, Cbc developers!

It seems Cbc is one of the very few opensource MILP solvers which
support parallelism, and it does a good job at it.

Sebastian