Are there any asm instructions that can speed up computation of min/max of vector of doubles/integers on Core i7 architecture?
Update:
I didn\'t
In response to your second question, it may be worthwhile to you to think about the way you collect and store this data.
You may store the data in a B-tree that keeps the data sorted at all times, requiring only logarithmic compare operations.
Then you know at all times where the maximum is.
http://en.wikipedia.org/wiki/B_tree