1: \begin{abstract}

2:   We present and analyze for a scalar linear evolution model problem a

3:   time multigrid algorithm for DG-discretizations in time. We derive

4:   asymptotically optimized parameters for the smoother, and also an

5:   asymptotically sharp convergence estimate for the two grid cycle.

6:   Our results hold for any A-stable time stepping scheme and represent

7:   the core component for space-time multigrid methods for parabolic

8:   partial differential equations.  Our time multigrid method has

9:   excellent strong and weak scaling properties for parallelization in

10:   time, which we show with numerical experiments.

11: \end{abstract}