Parallel receive transformation

[afiffon]

[afiffon]

Updating: This seems to have a lower impact on performance for now, since #63 resolved the main performance issue with AES.
It did seem that CPU was reaching single-thread bottleneck (never passing 100%, usually very high), but since changing this to use the armv8 AES instruction, it has become a non-issue! Might consider it in the future, or when need to compress data but throughput needs to be improved due to single-core performance limit.