BF1CVTL, BF1CVTL2, BF2CVTL, BF2CVTL2

8-bit floating-point convert to BFloat16 (vector). This instruction converts each 8-bit floating-point element from the lower or upper half of the source vector to BFloat16 while downscaling the value, and places the results in the 16-bit elements of the destination vector. BF1CVTL and BF2CVTL convert the elements from the lower half of the source vector while scaling the values by 2^{-UInt(FPMR.LSCALE[5:0])} and 2^{-UInt(FPMR.LSCALE2[5:0])}, respectively. BF1CVTL2 and BF2CVTL2 convert the elements from the upper half of the source vector while scaling the values by 2^{-UInt(FPMR.LSCALE[5:0])} and 2^{-UInt(FPMR.LSCALE2[5:0])}, respectively.

The 8-bit floating-point encoding format for BF1CVTL and BF1CVTL2 is selected by FPMR.F8S1. The 8-bit floating-point encoding format for BF2CVTL and BF2CVTL2 is selected by FPMR.F8S2.

Advanced SIMD
(FEAT_FP8)

if !IsFeatureImplemented(FEAT_FP8) then UNDEFINED; integer n = UInt(Rn); integer d = UInt(Rd); integer part = UInt(Q); constant integer elements = 64 DIV 8; boolean issrc2 = size == '11';