MPI source code: all the nodes broadcast a message, the zero nod receives messages. Nonblocking sending.

SUMMA并行矩阵的另一种MPI实现，在广播时以块为单位传输，运行速度明显提高-SUMMA parallel matrix another MPI implementation, the broadcast is transmitted in units of blocks, running speed is significantly improved

This is a common source of confusion for people new to MPI. You don't use MPI_Recv() to receive data sent by a broadcast; you use MPI_Bcast().