Name the three steps in the fetch-decode-execute loop.	Fetch, Decode, Execute.
What does the program counter hold?	The memory address of the next instruction to execute.
What is the program counter called on x86? On ARM?	rip on x86; pc on ARM.
By how much does the PC normally advance after each instruction?	By 4 bytes (the size of a typical instruction). For branches, it advances to the branch target instead.
How many bytes is a typical instruction?	4 bytes on ARM and most modern ISAs; 1–15 bytes (variable) on x86.
How many cycles does it take to fetch an instruction from the L1 instruction cache?	About 4 cycles.
How many cycles does a DRAM fetch take when the instruction is not in any cache?	Around 300 cycles.
What are µops (micro-operations)?	Simpler internal operations into which x86 CISC instructions are translated during decode, allowing the pipelined core to handle them uniformly.
What is the von Neumann architecture’s key insight?	Code and data both live in the same memory as bytes; the CPU reads and acts on bytes without distinguishing their origin.
How many instructions per second does a single CPU core execute?	About 3 billion per second.