关于如何只取得每组数据中的一个最大值的问题？

create table topic
(
Topicid int,
title nvarchar(10),
boardid int,
addtime datetime
)
insert topic select
1, 'abc', 100, '2007-1-1'
union all select 2, 'era', 101, '2007-1-2'
union all select 3, 'avx', 102, '2007-1-3'
union all select 4, 'zcv', 100, '2007-1-4'
union all select 5, 'jhv', 100, '2007-1-5'
union all select 6, 'ztw', 103, '2007-1-6'
union all select 7, 'xcv', 102, '2007-1-7'
union all select 8, 'zww', 104, '2007-1-8'
union all select 9, 'zqw', 105, '2007-1-9'
union all select 10, 'zti', 103, '2007-1-10'

--取得boardid相同的每组数据日期最大值：

select top 5 *
from topic as tmp
where not exists(select * from topic where boardid=tmp.boardid and addtime>tmp.addtime) order by addtime desc
--result 相同boardid,addtime最大的数据
Topicid title boardid addtime
----------- ---------- ----------- ------------------------------------------------------
10 zti 103 2007-01-10 00:00:00.000
9 zqw 105 2007-01-09 00:00:00.000
8 zww 104 2007-01-08 00:00:00.000
7 xcv 102 2007-01-07 00:00:00.000
5 jhv 100 2007-01-05 00:00:00.000

（所影响的行数为 5 行）

select a.* from topic a,
(
select boardid,max(addtime)as addtime from topic group by boardid
)b
where a.boardid=b.boardid and a.addtime=b.addtime
order by a.boardid

--result 相同boardid,addtime最大的数据
Topicid title boardid addtime

5 jhv 100 2007-01-05 00:00:00.000
2 era 101 2007-01-02 00:00:00.000
7 xcv 102 2007-01-07 00:00:00.000
10 zti 103 2007-01-10 00:00:00.000
8 zww 104 2007-01-08 00:00:00.000
9 zqw 105 2007-01-09 00:00:00.000