Given the memory latency impact on in-order cores, I wonder how much of the gap is core design and how much is Apple's tending to fit much larger caches to their designs.