SQL Server 查询不重复id的数据记录

想在这个下面示意的表中（这里的autoid唯一），
AUTOID ID AA BB Time
1 0001 35 20 2008102311
2 0001 65 40 2008102315
3 0001 89 56 2008102323
4 0002 26 55 2008102311
5 0002 104 69 2008102322
6 0002 106 36 2008101502
7 0002 106 38 2008101313
8 0003 56 265 2008101312
结果:
AUTOID ID AA BB Time
3 0001 89 56 2008102323
7 0002 106 36 2008101502
8 0003 56 265 2008101312

方法1：select * from tb t where not exists(select 1 from tb where id=t.id and aa>t.aa)

方法2：select * from tb a where a.autoid in (select max(autoid) as autoid from tb b group by b.id)

经过本人测试通过，如果问题请留言