0 - 1 of 1 tags for join plan

The way I thought Teradata always work is trying to keep things "local" to amp. Therefore, if A Join B on column C , the table A and Table B will be retrieved with not null on C and redistribute on hashcode of C and then join , do I under stand this right? (well of course, there are situations that you like to do duplicate to all amps in case one table is too small). The idea is that table needs to be distributed by hash of equal join column and then the data will resides on the same amp and they can perform join operation.