Those numbers aren’t directly comparable — caches usually trade off size for performance, and Apple’s L1 sits in performance somewhere between Ryzen’s L1 and L2

Those numbers aren’t directly comparable — caches usually trade off size for performance, and Apple’s L1 sits in performance somewhere between Ryzen’s L1 and L2. Ditto for Apple’s L2 versus Ryzen’s L2 and L3. I can’t remember the exact article I read that in, but it’s pretty obvious when looking at Anandtech’s latency charts. M1 latency: https://www.anandtech.com/show/17024/apple-m1-max-performance-review/2 https://www.anandtech.com/show/17024/apple-m1-max-performanc... Ryzen latency: https://www.anandtech.com/show/16214/amd-zen-3-ryzen-deep-dive-review-5950x-5900x-5800x-and-5700x-tested/5 https://www.anandtech.com/show/16214/amd-zen-3-ryzen-deep-di...