Steve,

Both DISTINCT and GROUP BY require that the entire record set be retreived and then compared to each other. There is no way to optimize those operations.

If you can describe the goal, perhaps there is another way to get the result you want.