---

General API

• Memory function

• Float16 related functions

• Defines provided by NSIMD

• NSIMD pack and related functions

• NSIMD C++20 concepts

SIMD operators

Miscellaneous

• Vector length (len)

• Value broadcast (set1)

• Logical value broadcast (set1l)

• Horizontal sum (addv)

• Blend (if_else1)

• Check all elements (all)

• Check for one true elements (any)

• Count true elements (nbtrue)

• Fill vector with increasing values (iota)

• Build mask for ending loops (mask_for_loop_tail)

Loads & stores

• Loadu (loadu)

• MaskoLoadu1 (masko_loadu1)

• MaskzLoadu1 (maskz_loadu1)

• Load array of structure (load2u)

• Load array of structure (load3u)

• Load array of structure (load4u)

• Loada (loada)

• MaskoLoada (masko_loada1)

• MaskzLoada (maskz_loada1)

• Load array of structure (load2a)

• Load array of structure (load3a)

• Load array of structure (load4a)

• Load vector of logicals (loadlu)

• Load vector of logicals (loadla)

• Storeu (storeu)

• MaskStoreu1 (mask_storeu1)

• Store2u (store2u)

• Store into array of structures (store3u)

• Store into array of structures (store4u)

• Storea (storea)

• MaskStorea1 (mask_storea1)

• Store into array of structures (store2a)

• Store into array of structures (store3a)

• Store into array of structures (store4a)

• Gather elements from memory into a SIMD vector (gather)

• Gather elements from memory into a SIMD vector (gather_linear)

• Scatter elements from SIMD vector to memory (scatter)

• Scatter elements from SIMD vector to memory (scatter_linear)

• Store vector of logicals (storelu)

• Store vector of logicals (storela)

Bits manipulation operators

• Bitwise or (orb)

• Bitwise and (andb)

• Bitwise andnot (andnotb)

• Bitwise not (notb)

• Bitwise xor (xorb)

• Right shift in zeros (shr)

• Left shift (shl)

• Arithmetic right shift (shra)

Logicals operators

• Logical or (orl)

• Logical and (andl)

• Logical andnot (andnotl)

• Logical xor (xorl)

• Logical not (notl)

• Build mask from logicals (to_mask)

• Build logicals from data (to_logical)

Basic arithmetic operators

• Addition (add)

• Subtraction (sub)

• Multiplication (mul)

• Division (div)

• Opposite (neg)

• Minimum (min)

• Maximum (max)

• Absolute value (abs)

• Fused multiply-add (fma)

• Fused negate-multiply-add (fnma)

• Fused multiply-substract (fms)

• Fused negate-multiply-substract (fnms)

• Reciprocal (rec)

• Reciprocal with relative error at most \(2^{-11}\) (rec11)

• Reciprocal with relative error at most \(2^{-8}\) (rec8)

• Square root (sqrt)

• Square root with relative error at most \(2^{-11}\) (rsqrt11)

• Square root with relative error at most \(2^{-8}\) (rsqrt8)

• Addition using saturation (adds)

• Subtraction using saturation (subs)

• Cube root (cbrt_u35)

• Cube root (cbrt_u10)

• Euclidean distance (hypot_u05)

• Euclidean distance (hypot_u35)

• Floating-point remainder (remainder)

• Floating-point remainder (fmod)

Comparison operators

• Compare for equality (eq)

• Compare for inequality (ne)

• Compare for greater-than (gt)

• Compare for greater-or-equal-than (ge)

• Compare for lesser-than (lt)

• Compare for lesser-or-equal-than (le)

Rounding functions

• Rounding up to integer value (ceil)

• Rounding down to integer value (floor)

• Rounding towards zero to integer value (trunc)

• Rounding to nearest integer value, tie to even (round_to_even)

Conversion operators

• Reinterpret vector (reinterpret)

• Reinterpret vector of logicals (reinterpretl)

• Convert vector (cvt)

• Convert vector to larger type (upcvt)

• Convert vector to narrow type (downcvt)

Shuffle functions

• Zip low halves (ziplo)

• Zip high halves (ziphi)

• Unziplo (unziplo)

• Unziphi (unziphi)

• Zip (zip)

• Unzip (unzip)

Trigonometric functions

• Sine (sin_u35)

• Cosine (cos_u35)

• Tangent (tan_u35)

• Arcsine (asin_u35)

• Arccosine (acos_u35)

• Arctangent (atan_u35)

• Arctangent (atan2_u35)

• Sine (sin_u10)

• Cosine (cos_u10)

• Tangent (tan_u10)

• Arcsine (asin_u10)

• Arccosine (acos_u10)

• Arctangent (atan_u10)

• Arctangent (atan2_u10)

• Sine of pi times argument (sinpi_u05)

• Cosine of pi times argument (cospi_u05)

Exponential and logarithmic functions

• Natural logarithm (log_u35)

• Natural logarithm (log_u10)

• Base-e exponential (exp_u10)

• Power (pow_u10)

• Base-2 exponential (exp2_u10)

• Base-2 exponential (exp2_u35)

• Base-10 exponential (exp10_u10)

• Base-10 exponential (exp10_u35)

• Exponential minus 1 (expm1_u10)

• Base-10 logarithm (log10_u10)

• Base-2 logarithm (log2_u10)

• Base-2 logarithm (log2_u35)

• Logarithm of 1 plus argument (log1p_u10)

• Log gamma (lgamma_u10)

• True gamma (tgamma_u10)

• Complementary error (erf_u10)

• Complementary error (erfc_u15)

Hyperbolic functions

• Hyperbolic sine (sinh_u10)

• Hyperbolic cosine (cosh_u10)

• Hyperbolic tangent (tanh_u10)

• Hyperbolic sine (sinh_u35)

• Hyperbolic cosine (cosh_u35)

• Hyperbolic tangent (tanh_u35)

• Inverse hyperbolic sine (asinh_u10)

• Inverse hyperbolic cosine (acosh_u10)

• Inverse hyperbolic tangent (atanh_u10)