SVE Instruction List by Dougall Johnson
FMAXNMQV: Floating-point maximum number recursive reduction of quadword vector segments
FMAXNMQV Vd.8H, Pg, Zn.H (SVE2.1 (SME2.1
128-bit SVE
Take the maximum number (i.e. non-QNaN) across corresponding active 16-bit float elements from each 128-bit segment of (1), storing the result in (3). For any element where the corresponding predicate bit in (2) is 0, interpret it as QNaN.
256-bit SVE
Take the maximum number (i.e. non-QNaN) across corresponding active 16-bit float elements from each 128-bit segment of (1), storing the result in (3). For any element where the corresponding predicate bit in (2) is 0, interpret it as QNaN.
512-bit SVE
Take the maximum number (i.e. non-QNaN) across corresponding active 16-bit float elements from each 128-bit segment of (1), storing the result in (3). For any element where the corresponding predicate bit in (2) is 0, interpret it as QNaN.
Larger sizes
1024-bit SVE
Take the maximum number (i.e. non-QNaN) across corresponding active 16-bit float elements from each 128-bit segment of (1), storing the result in (3). For any element where the corresponding predicate bit in (2) is 0, interpret it as QNaN.
2048-bit SVE
Take the maximum number (i.e. non-QNaN) across corresponding active 16-bit float elements from each 128-bit segment of (1), storing the result in (3). For any element where the corresponding predicate bit in (2) is 0, interpret it as QNaN.
Switch to High-DPISwitch to Low-DPI
Report mistakes or give feedback
Inspired by and based on the x86/x64 SIMD Instruction List by Daytime.