is hive have its own map reduce program?

i want to implement hive+hadoop map reduce program on my aplication, i still wondering,because i have try many times about query and finding information about map reduce program in hive..

my question is,is hive have its own map reduce program? because when i try a litle bit complex query the log is like :

Job running in-process (local Hadoop)
2011-05-26 14:10:02,004 null map = 100%,  reduce = 100%
Ended Job = job_local_0001

there is map reduce process in that query,and then if i try to google some of map reduce program implementing in hive,there also have alternative to create our map reduce program..

my opinion is,we can use hive to process data mining(large queries)(i dont know right or wrong?please correct me) we can create our map reduce program to implement in hive for spesific task(is that true?)

is there any explanation about this?

thanks for you all explanantion and attention,and sorry for my bad english..thanks..:D

No it uses Hadoop's MapReduce. But it uses MapReduce to aggregate the data that is needed for your query. So if you've searched for every value so a specific key, Hive will run a MapReduce job on your input to get the data for you.

Basically HIVE uses mapreduce and transforms the SQL-Like language to MapReduce jobs.