How parallel hint works on the full table scan?

I have couple of interview questions. Can any one please through some lights on this.

1. What are the init parameters to improve the full tablespace in SQL QUERY?

2. I have table which has huge number of records. Let us say, table has 500 million records. My business logic is forcing me to scan the entire table. I am using PARALLEL hint to get better performance. How the parrellel process works internally? How multiple process works on the same table and join the results?? Since there is only one table and how multiple process hit the same table and join the result?

1. what is a full tablespace scan?
2. tahiti.oracle.com

1. Sorry. It is my tying mistake. It is Full table scan. Not full tablespace.
2. What is this? tahiti.oracle.com

think about partitioning maybe, if you can split it on what you are searching on.

Full scanning 500 million rows will take a long time - why *MUST* you full scan it and not index the relevant columns?

Dav, I am not getting answer for my question.

1. What are INIT parameters for improving the full tablescan?
2. How does PARALLEL process works for full tablescan on non partitioned table?

These two are the interview questions. I am trying to get answer for these.

1.DB Files Multiblock Read Count
2 Read concepts as suggested by marist