F1CVTL, F1CVTL2, F2CVTL, F2CVTL2

8-bit floating-point convert to half-precision (vector)

This instruction converts each 8-bit floating-point element from the lower or upper half of the source vector to half-precision while downscaling the value, and places the results in the 16-bit elements of the destination vector. F1CVTL and F2CVTL convert the elements from the lower half of the source vector while scaling the values by 2^{-UInt(FPMR.LSCALE[3:0])} and 2^{-UInt(FPMR.LSCALE2[3:0])}, respectively. F1CVTL2 and F2CVTL2 convert the elements from the upper half of the source vector while scaling the values by 2^{-UInt(FPMR.LSCALE[3:0])} and 2^{-UInt(FPMR.LSCALE2[3:0])}, respectively.

The 8-bit floating-point encoding format for F1CVTL and F1CVTL2 is selected by FPMR.F8S1. The 8-bit floating-point encoding format for F2CVTL and F2CVTL2 is selected by FPMR.F8S2.

Advanced SIMD
(FEAT_FP8)

Decode for all variants of this encoding

if !IsFeatureImplemented(FEAT_FP8) then EndOfDecode(Decode_UNDEF); constant integer n = UInt(Rn); constant integer d = UInt(Rd); constant integer part = UInt(Q); constant integer elements = 64 DIV 8; constant boolean issrc2 = size == '01';