That's why we are focusing on adding GPU cores instead, but they can only do the same operation on a lot of data in parallel (much like mass transit).