Counting Kmers for Biological Sequences at Large Scale