Oh a brief followup question if you have interest -- How does one know when it's time to design your own chip? Versus take something off the shelf? At what level of company or product maturity is this realization even likely to be discovered?