The NCC$2000$ processor provides a cache with cache lines that are $16$ bytes

long. Suppose the cache is empty when the processor starts running a

program. Using a loop, this program writes a byte to a memory address

held in the $\%$rax register and then increments the address in the $\%$rax

register by $32.$ Suppose this loop iterates $64$ times on a machine with

$4$ GB of physical memory.

How many cache misses does this loop incur? Ignore cache misses for

instruction fetches.