官方的参考文档
http://technet.microsoft.com/zh-cn/library/ms187074%28SQL.90%29.aspx
他们作用于比较运算符和子查询之间，作用类似Exists、not exists、in、not in以及其他逻辑意义，这些语法同样被SQLServer2000支持但是很少看到有人用它们。
set nocount on
use tempdb
go
if (object_id ('t1' ) is not null ) drop table t1
create table t1 (n int )
insert into t1 select 2 union select 3
if (object_id ('t2' ) is not null ) drop table t2
create table t2 (n int )
insert into t2 select 1 union select 2 union select 3 union select 4
select * from t2 where n> all (select n from t1 ) --4
select * from t2 where n> any (select n from t1 ) --3,4
--select * from t2 where n>some(select n from t1) --3,4
select * from t2 where n= all (select n from t1 ) --无数据
select * from t2 where n= any (select n from t1 ) --2,3
--select * from t2 where n=some(select n from t1) --2,3
select * from t2 where n< all (select n from t1 ) --1
select * from t2 where n< any (select n from t1 ) --1,2
--select * from t2 where n<some(select n from t1) --1,2
select * from t2 where n<> all (select n from t1 ) --1,4
select * from t2 where n<> any (select n from t1 ) --1,2,3,4
--select * from t2 where n<>some(select n from t1)--1,2,3,4
set nocount off
注意，如果t1中包含null数据，那么所有All相关的比较运算将不会返回任何结果，原因就不用多解释了。而因为t1和t2表的null的存在他们和not exists之类的比较符会有一些区别。
比如下面两句
select * from t2 a where not exists(select 1 from t1 where n>=a.n)
select * from t2 where n >all(select n from t1)
他们逻辑上意义很像但是对于null的处理却是恰恰相反，第一句会忽略子查询的null而把t2的null同时查出来，第二句却是忽略了t2的null同时会因为t1中的null而无法查询到数据。