Remove the duplicate rows but keep the most recent and oldest records

Question

I have a Product table, some products have duplicate rows(same productID). I want to keep the most recent and the oldest records based on ETL load date (or min and max productKey), delete others. How to do this in a simple way?
ProductKey | ProductID |ETL_Load_Date
1001 | 501 | 2014-01-01
1002 | 501 | 2014-01-02
1003 | 501 | 2014-01-03
1011 | 502 | 2014-01-01
1012 | 502 | 2014-01-02
1013 | 502 | 2014-01-03

Answer 1

declare @table table (ProductKey int, ProductID int, ETL_Load_Date date)
insert @table (ProductKey, ProductID, ETL_Load_Date)
values
( 1001 , 501 , '2014-01-01' ), ( 1002 , 501 , '2014-01-02' ),
( 1003 , 501 , '2014-01-03' ), ( 1011 , 502 , '2014-01-01' ),
( 1012 , 502 , '2014-01-02' ), ( 1013 , 502 , '2014-01-03' )
SELECT *
FROM @table t
INNER JOIN (
SELECT MIN(etl_load_date) min_load_date, MAX(etl_load_date) max_load_date, ProductID
FROM @table
GROUP BY ProductID
) m
ON t.ProductID = m.ProductID
AND t.ETL_Load_Date IN (min_load_date,max_load_date)
This will show both the min and max load dates, for each ProductID
Thanks to Saeid for posting some test objects.