RX 580 is slower than HD530 when encoding HEVC

SMBIOS: iMac17,1. I did a quick test about HEVC encoding speed in FCPX

Round 1: Compressor uses ~75% of HD530, total time 1min43s.

Round 2: Add AAPL,slot-name=Slot-1. Compressor uses only ~25% of RX580, total time 1min46s.

Just wondering is AMD being too slow or is Intel being too fast?