> but what you can you do.

Not make sweeping generalizations based on one benchmark you didn't write?