how small to split a table?

I've got one logging table that is over 330 million rows to store 6
months' worth of data. It consists of two integers and a 4-character
long string. I have one primary key which is the two integers, and
an additional index on the second integer.

I'm planning to use inheritance to split the table into a bunch of
smaller ones by using a modulo function on one of the integers on
which we scan often.

My question is how small to make each inherited piece? If I do
modulo 10, then each sub-table will be between 32 and 34 million rows
today based on current distribution.

If I expect to increase traffic 2 times over the next year (thus
doubling my logs) what would you recommend?