FMAXNMV: Floating-point maximum number recursive reduction to scalar

FMAXNMV Hd, Pg, Zn.H (SVE (SME

float16_t svmaxnmv[_f16](svbool_t pg, svfloat16_t op)

128-bit SVE

Find the maximum number (i.e. non-QNaN) of active 16-bit float elements from (1), using a tree reduction, storing the result in (2). Inactive elements are interpreted as QNaN.

256-bit SVE

Find the maximum number (i.e. non-QNaN) of active 16-bit float elements from (1), using a tree reduction, storing the result in (2). Inactive elements are interpreted as QNaN.

512-bit SVE

Find the maximum number (i.e. non-QNaN) of active 16-bit float elements from (1), using a tree reduction, storing the result in (2). Inactive elements are interpreted as QNaN.

Larger sizes

1024-bit SVE

Find the maximum number (i.e. non-QNaN) of active 16-bit float elements from (1), using a tree reduction, storing the result in (2). Inactive elements are interpreted as QNaN.

2048-bit SVE

Find the maximum number (i.e. non-QNaN) of active 16-bit float elements from (1), using a tree reduction, storing the result in (2). Inactive elements are interpreted as QNaN.

Report mistakes or give feedback
Inspired by and based on the x86/x64 SIMD Instruction List by Daytime.