Memory's Pyramid

10 atomic recall cards. Export to Anki and let spaced repetition do its slow work.

In Anki: File → Import, choose this TSV, set field separator to Tab, deck = Sand to Silicon · Ch 24, note type = Basic.

Front	Back
What is the approximate latency of a register access?	~1 cycle.
What is the approximate latency of an L1 cache access?	~3–4 cycles.
What is the approximate latency of an L3 cache access?	~40 cycles.
What is the approximate latency of a DRAM access?	~300 cycles — the 300× cliff the pyramid exists to hide.
What size range does L3 cache span, and what qualifies the two ends?	8–64 MB: consumer CPUs at the low end (8–32 MB), server-class parts at the high end.
What is a cache line, and how big is it on x86?	The unit of data movement between cache levels — 64 bytes on x86 and most ARM; 128 bytes on Apple Silicon.
What is temporal locality?	The tendency of programs to reuse recently accessed memory locations soon after accessing them.
What is spatial locality?	The tendency of programs to access memory addresses near recently accessed ones.
What does the TLB cache, and what is a typical L1 TLB size?	Virtual-to-physical address translations. The L1 TLB often holds around sixty-four entries.
What is the memory wall?	The growing gap between CPU speed and DRAM speed — documented by Hennessy and Patterson as roughly doubling every few years through the 1990s.