Short version:
The best/fastest way to load large amounts of data from a comma delimited text file into an SQL Server table. Where the text file contains date fields in ccyy/mm/dd format and the SQL Server table defines those fields as datetime data types.

Details:
When I attempt to load files (using either bcp or BULK INSERT) containing datetime data the load process errors because the datetime fields in my text file are in ccyy/mm/dd format and the default format for SQL Server is mm/dd/yy. I have been unable to change the default format by using the SET DATEFORMAT statement (apparently the SET DATEFORMAT statement will not work for bcp because bcp runs outside of the SQL Server session???).
The only alternatives that I have come up with are: 1) Change the format of date fields in the text file from ccyy/mm/dd to mm/dd/ccyy. 2) Create a temporary table that defines the date fields as a char(n) datatype. Then load the data into the temp table. Then SET the DATEFORMAT to ccyy/mm/dd. Then copy the temp table into the permanent table (the permanent table using datetime data types).

Both of these alternatives would require additional processing time. Since this is a process that loads large amounts of data on a monthly (soon to be weekly) basis, speed is of the essence.

I would appreciate any suggestions.

Thanks!