Why is Apple Rosetta 2 fast? (2022)

Bit-twiddling abstract addition with unknown bits (2020)

Why is Rosetta 2 fast?

Reading bits with zero refill latency

Faster zlib/DEFLATE decompression on the Apple M1 (and x86)

Parallelising Huffman decoding and x86 disassembly by synchronising prefix codes

Bitwise conversion of doubles using only floating-point multiplication and addition

Faster CRC32 on the Apple M1

Apple M1: Load and Store Queue Measurements

Bitwise conversion of doubles using only floating point multiply and divide

Bit-Twiddling: Addition with Unknown Bits

Exploiting Dolphin - Part 1