hibernate vs JDBC

We have a scnario where in once a week a batch program will run and it will process 2 million records. we have to retrieve them from a a file, store them in a table and process the records and update the table.

For the database part, we are deciding whether to use the a OR mapping tool like Hibernate or use the plain JDBC queries. There are nearly 40 tables. Our point of concern is that Hibernate may take more time and memory for processing and since this is a large amount of data, we may have a performance hit than using plaing JDBC calls for retrieval and updates. Is that assumption correct? Will there be a performance hit?

you didn't mention the size of the record. 2 mil doesn't sound that much. Hibernate should be ok.

This question would be better off in the JDBC forum...

If you are still going to create the JavaBeans that Hibernate would require even if you still use JDBC, the biggest hit you may encounter is at startup when Hibernate reads all it's config data and starts your session factory. Beyond that it will likely be the same either way.