I would guess that you're using different network libraries on the two OSes and that's responsible for the difference. Different settings will be used depending on your network library, and that can make a difference to query execution. Since you have already run a trace, have a look through it again, and (depending on what you set it to capture), you should have a row like this:

SET ROWCOUNT 0 SET TEXTSIZE 2147483647 SET NOCOUNT OFF SET CONCAT_NULL_YIELDS_NULL ON SET ARITHABORT ON SET LOCK_TIMEOUT -1 SET QUERY_GOVERNOR_COST_LIMIT 0 SET DEADLOCK_PRIORITY NORMAL SET TRANSACTION ISOLATION LEVEL READ COMMITTED SET ANSI_NULLS ON SET ANSI_NULL_DFLT_ON ON SET ANSI_PADDING ON SET ANSI_WARNINGS ON SET CURSOR_CLOSE_ON_COMMIT OFF SET IMPLICIT_TRANSACTIONS OFF SET QUOTED_IDENTIFIER ON

Compare this for the slow client and the fast client. Are any of the settings different?

John