Abstract
We present some work in progress on lower bounds for information theoretic MPC. One set of results is about the number of messages required to compute non-trivial functions such the and, we show that a quadratic number of messages in the number of players is required. In the second line of work we show that protocols following the traditional secret-sharing based design paradigm cannot be significantly improved wrt round and communication complexity, even with preprocessing.