Efficiently selecting "sub maxes" from SQL Table

Apologies for the title, I am unsure how to 开发者_开发百科express my problem... I have a table that looks something like this:

A     | B
--------------
01    | 1
01A   | 1
01B   | 1
01C   | 1
02    | 1
03    | 1
03    | 2
03A   | 1
03B   | 1

Sometimes column A is qualified with a letter (##A,##B), sometimes it is not (##). I would need to

SELECT A,B from Table

if there were no suffixes. With the suffix, however, I need to select the maximum value of A for each B. (That was phrased poorly). I would expect the following results for the table above:

A     | B
--------------
01C   | 1
02    | 1
03    | 2
03B   | 1

Right now, I am using a subselect and my query is:

SELECT A,B FROM Table t1
WHERE t1.A = (SELECT MAX(t2.A) FROM Table t2WHERE LEFT(t1.A,2) = LEFT(t2.A,2)

This is incredibly inefficient, however, and my table is pretty large, so they query is taking far too long to run. Is there a better way?

Thanks!

;with cteRowNumbers as (
    select A, B, ROW_NUMBER() over(partition by B, left(A,2) order by A desc) as RowNumber
        from YourTable
 )
 select A, B
    from cteRowNumbers
    where RowNumber = 1
    order by A

If you find common table expressions unpleasant, there's this which is essentially the same thing as the CTE example.

SELECT V.* 

FROM (
    SELECT
     A
    ,B
    ,ROW_NUMBER()OVER(PARTITION BY B,LEFT(A,2) ORDER BY A DESC) AS RowNumber

    FROM YourTable
) AS V

WHERE V.RowNumber = 1