Largely because a lot of compression is contextual to the data around it. That makes most decompression a poor fit for parallelism. The primary choices are spa

Largely because a lot of compression is contextual to the data around it. That makes most decompression a poor fit for parallelism. The primary choices are space savings and power efficiency. Dedicated hardware and serial decoding/encoding often win out as a result.