is this execution (inside a trans block):
insert into table_1 (nam1, nam2, ..., namN)
values(val11, val12, ..., val1N);
insert into table_2 (nam1, nam2, ..., namN)
values(val21, val22, ..., val2N);
...
insert into table_M (namM1, namM2, ..., namMN)
values(valM1, valM2, ..., valMN)
MUCH faster than if I do it one at a time?