Intrinsics allowing you to perform compare and select operations on all types of vectors. More...

Overview

Intrinsics allowing you to perform compare and select operations on all types of vectors.

sub_lt intrinsic example

v32int16 a                 = //init
v32int16 b                 = //init
v32int16 res_sub           = //init
unsigned long long res_cmp = //init
 
res_sub = sub_lt(a, b, res_cmp);

Functions
v16float	sel (v16float a, v16float b, unsigned int sel)
	Performs lane wise selection between vectors a and b. If a bit of sel if zero the lane of vector a is selected, else the lane of vector b is selected.

Vector min/max and combined intrinsics for datatype v64uint8
v64uint8	sub_lt (v64uint8 a, v64uint8 b, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64uint8	sub_lt (v64uint8 a, v64uint8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64uint8	sub_ge (v64uint8 a, v64uint8 b, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v64uint8	sub_ge (v64uint8 a, v64uint8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v64uint8	maxdiff_lt (v64uint8 a, v64uint8 b, unsigned long long &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v64int8	maxdiff_lt (v64int8 a, v64int8 b, unsigned long long &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v64uint8	maxdiff_lt (v64uint8 a, v64uint8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64uint8	maxdiff (v64uint8 a, v64uint8 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v64uint8	maxdiff (v64uint8 a, v64uint8 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v64uint8	min_ge (v64uint8 a, v64uint8 b, unsigned long long &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v64int8	min_ge (v64int8 a, v64int8 b, unsigned long long &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v64uint8	min_ge (v64uint8 a, v64uint8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64uint8	min (v64uint8 a, v64uint8 b)
	Calculates the minimum between two input vectors.

v64uint8	min (v64uint8 a, v64uint8 b, bool sgn)
	Calculates the minimum between two input vectors.

v64uint8	max_lt (v64uint8 a, v64uint8 b, unsigned long long &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v64int8	max_lt (v64int8 a, v64int8 b, unsigned long long &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v64uint8	max_lt (v64uint8 a, v64uint8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64uint8	max (v64uint8 a, v64uint8 b)
	Calculates the maximum between two input vectors.

v64uint8	max (v64uint8 a, v64uint8 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v64int8
v64int8	sub_lt (v64int8 a, v64int8 b, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64int8	sub_lt (v64int8 a, v64int8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v64int8	sub_ge (v64int8 a, v64int8 b, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v64int8	sub_ge (v64int8 a, v64int8 b, bool sgn, unsigned long long &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v64int8	maxdiff_lt (v64int8 a, v64int8 b, bool sgn, unsigned long long &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v64int8	maxdiff (v64int8 a, v64int8 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v64int8	maxdiff (v64int8 a, v64int8 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v64int8	min_ge (v64int8 a, v64int8 b, bool sgn, unsigned long long &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v64int8	min (v64int8 a, v64int8 b)
	Calculates the minimum between two input vectors.

v64int8	min (v64int8 a, v64int8 b, bool sgn)
	Calculates the minimum between two input vectors.

v64int8	max_lt (v64int8 a, v64int8 b, bool sgn, unsigned long long &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v64int8	max (v64int8 a, v64int8 b)
	Calculates the maximum between two input vectors.

v64int8	max (v64int8 a, v64int8 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v32uint16
v32uint16	sub_lt (v32uint16 a, v32uint16 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32uint16	sub_lt (v32uint16 a, v32uint16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32uint16	sub_ge (v32uint16 a, v32uint16 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v32uint16	sub_ge (v32uint16 a, v32uint16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v32uint16	maxdiff_lt (v32uint16 a, v32uint16 b, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v32int16	maxdiff_lt (v32int16 a, v32int16 b, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v32uint16	maxdiff_lt (v32uint16 a, v32uint16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32uint16	maxdiff (v32uint16 a, v32uint16 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v32uint16	maxdiff (v32uint16 a, v32uint16 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v32uint16	min_ge (v32uint16 a, v32uint16 b, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v32int16	min_ge (v32int16 a, v32int16 b, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v32uint16	min_ge (v32uint16 a, v32uint16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32uint16	min (v32uint16 a, v32uint16 b)
	Calculates the minimum between two input vectors.

v32uint16	min (v32uint16 a, v32uint16 b, bool sgn)
	Calculates the minimum between two input vectors.

v32uint16	max_lt (v32uint16 a, v32uint16 b, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v32int16	max_lt (v32int16 a, v32int16 b, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v32uint16	max_lt (v32uint16 a, v32uint16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32uint16	max (v32uint16 a, v32uint16 b)
	Calculates the maximum between two input vectors.

v32uint16	max (v32uint16 a, v32uint16 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v32int16
v32int16	sub_lt (v32int16 a, v32int16 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32int16	sub_lt (v32int16 a, v32int16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v32int16	sub_ge (v32int16 a, v32int16 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v32int16	sub_ge (v32int16 a, v32int16 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v32int16	maxdiff_lt (v32int16 a, v32int16 b, bool sgn, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v32int16	maxdiff (v32int16 a, v32int16 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v32int16	maxdiff (v32int16 a, v32int16 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v32int16	min_ge (v32int16 a, v32int16 b, bool sgn, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v32int16	min (v32int16 a, v32int16 b)
	Calculates the minimum between two input vectors.

v32int16	min (v32int16 a, v32int16 b, bool sgn)
	Calculates the minimum between two input vectors.

v32int16	max_lt (v32int16 a, v32int16 b, bool sgn, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v32int16	max (v32int16 a, v32int16 b)
	Calculates the maximum between two input vectors.

v32int16	max (v32int16 a, v32int16 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v16uint32
v16uint32	sub_lt (v16uint32 a, v16uint32 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16uint32	sub_lt (v16uint32 a, v16uint32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16uint32	sub_ge (v16uint32 a, v16uint32 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v16uint32	sub_ge (v16uint32 a, v16uint32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v16uint32	maxdiff_lt (v16uint32 a, v16uint32 b, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v16int32	maxdiff_lt (v16int32 a, v16int32 b, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v16uint32	maxdiff_lt (v16uint32 a, v16uint32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16uint32	maxdiff (v16uint32 a, v16uint32 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v16uint32	maxdiff (v16uint32 a, v16uint32 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v16uint32	min_ge (v16uint32 a, v16uint32 b, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v16int32	min_ge (v16int32 a, v16int32 b, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v16uint32	min_ge (v16uint32 a, v16uint32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16uint32	min (v16uint32 a, v16uint32 b)
	Calculates the minimum between two input vectors.

v16uint32	min (v16uint32 a, v16uint32 b, bool sgn)
	Calculates the minimum between two input vectors.

v16uint32	max_lt (v16uint32 a, v16uint32 b, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v16int32	max_lt (v16int32 a, v16int32 b, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v16uint32	max_lt (v16uint32 a, v16uint32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16uint32	max (v16uint32 a, v16uint32 b)
	Calculates the maximum between two input vectors.

v16uint32	max (v16uint32 a, v16uint32 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v16int32
v16int32	sub_lt (v16int32 a, v16int32 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16int32	sub_lt (v16int32 a, v16int32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

v16int32	sub_ge (v16int32 a, v16int32 b, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v16int32	sub_ge (v16int32 a, v16int32 b, bool sgn, unsigned int &cmp)
	Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

v16int32	maxdiff_lt (v16int32 a, v16int32 b, bool sgn, unsigned int &cmp)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

v16int32	maxdiff (v16int32 a, v16int32 b)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v16int32	maxdiff (v16int32 a, v16int32 b, bool sgn)
	Calculates the difference between two input vectors and returns the maximum between the difference and zero.

v16int32	min_ge (v16int32 a, v16int32 b, bool sgn, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v16int32	min (v16int32 a, v16int32 b)
	Calculates the minimum between two input vectors.

v16int32	min (v16int32 a, v16int32 b, bool sgn)
	Calculates the minimum between two input vectors.

v16int32	max_lt (v16int32 a, v16int32 b, bool sgn, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v16int32	max (v16int32 a, v16int32 b)
	Calculates the maximum between two input vectors.

v16int32	max (v16int32 a, v16int32 b, bool sgn)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v32bfloat16
v32bfloat16	min_ge (v32bfloat16 a, v32bfloat16 b, unsigned int &cmp)
	Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

v32bfloat16	min (v32bfloat16 a, v32bfloat16 b)
	Calculates the minimum between two input vectors.

v32bfloat16	max_lt (v32bfloat16 a, v32bfloat16 b, unsigned int &cmp)
	Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

v32bfloat16	max (v32bfloat16 a, v32bfloat16 b)
	Calculates the maximum between two input vectors.

Vector min/max and combined intrinsics for datatype v16float
unsigned int	lt (v16float v1, v16float v2)
	Vector comparison. Returns lane by lane comparison v1.lane < v2.lane.

v16float	max_lt (v16float v1, v16float v2, unsigned int &cmp)
	Vector comparison. Returns max(v1,v2) along with lane by lane comparison v1.lane0 < v2.lane0.

v16float	max (v16float v1, v16float v2)
	Vector comparison. Returns max(v1,v2)

unsigned int	ge (v16float v1, v16float v2)
	Vector comparison. Returns lane by lane comparison v1.lane >= v2.lane.

unsigned int	le (v16float a, v16float b)
	Checks for which vector lanes a <= b is fulfilled.

v16float	min_ge (v16float v1, v16float v2, unsigned int &cmp)
	Vector comparison. Returns min(v1,v2) along with lane by lane comparison v1.lane >= v2.lane.

v16float	min (v16float v1, v16float v2)
	Vector comparison. Returns min(v1,v2)

unsigned int	gt (v16float v1, v16float v2)
	Vector comparison. Returns lane by lane comparison v1.lane > v2.lane.

unsigned int	eq (v16float a, v16float b)
	Checks for which vector lanes a == b is fulfilled.

unsigned int	ne (v16float a, v16float b)
	Checks for which vector lanes a != b is fulfilled.

v16float	abs (v16float v1)
	absolute value of Vector. Returns abs(v1).

v16float	min_abs (v16float v1, v16float v2)
	Vector comparison.

v16float	max_abs (v16float v1, v16float v2)
	Vector comparison.

unsigned int	ge_abs (v16float v1, v16float v2)
	Vector comparison. Returns lane by lane comparison v1.lane >= abs(v2.lane)

unsigned int	lt_abs (v16float v1, v16float v2)

unsigned int	operator< (v16float a, v16float b)

unsigned int	operator>= (v16float a, v16float b)

unsigned int	operator> (v16float a, v16float b)

unsigned int	operator== (v16float a, v16float b)

unsigned int	operator!= (v16float a, v16float b)

Function Documentation

◆ abs()

v16float abs ( v16float v1 )

absolute value of Vector. Returns abs(v1).

Parameters

v1 Vector a

Returns: Result of operation abs(a)

◆ eq()

unsigned int eq	(	v16float	a,
		v16float	b
	)

Checks for which vector lanes a == b is fulfilled.

Parameters

a	Vector a
b	Vector b

Returns: Result of: a == b

◆ ge()

unsigned int ge	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns lane by lane comparison v1.lane >= v2.lane.

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation (a>=b)

◆ ge_abs()

unsigned int ge_abs	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns lane by lane comparison v1.lane >= abs(v2.lane)

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation (a>=abs(b))

◆ gt()

unsigned int gt	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns lane by lane comparison v1.lane > v2.lane.

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation (a>b)

◆ le()

unsigned int le	(	v16float	a,
		v16float	b
	)

Checks for which vector lanes a <= b is fulfilled.

Parameters

a	Vector a
b	Vector b

Returns: Result of: a <= b

◆ lt()

unsigned int lt	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns lane by lane comparison v1.lane < v2.lane.

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation (a<b)

◆ lt_abs()

unsigned int lt_abs	(	v16float	v1,
		v16float	v2
	)

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation (a

◆ max() [1/14]

v16float max	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns max(v1,v2)

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation max(a,b)

◆ max() [2/14]

v16int32 max	(	v16int32	a,
		v16int32	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [3/14]

v16int32 max	(	v16int32	a,
		v16int32	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max() [4/14]

v16uint32 max	(	v16uint32	a,
		v16uint32	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [5/14]

v16uint32 max	(	v16uint32	a,
		v16uint32	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max() [6/14]

v32bfloat16 max	(	v32bfloat16	a,
		v32bfloat16	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [7/14]

v32int16 max	(	v32int16	a,
		v32int16	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [8/14]

v32int16 max	(	v32int16	a,
		v32int16	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max() [9/14]

v32uint16 max	(	v32uint16	a,
		v32uint16	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [10/14]

v32uint16 max	(	v32uint16	a,
		v32uint16	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max() [11/14]

v64int8 max	(	v64int8	a,
		v64int8	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [12/14]

v64int8 max	(	v64int8	a,
		v64int8	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max() [13/14]

v64uint8 max	(	v64uint8	a,
		v64uint8	b
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a, b)

◆ max() [14/14]

v64uint8 max	(	v64uint8	a,
		v64uint8	b,
		bool	sgn
	)

Calculates the maximum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a, b)

◆ max_abs()

v16float max_abs	(	v16float	v1,
		v16float	v2
	)

Vector comparison.

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation max(a,abs(b))

◆ max_lt() [1/14]

v16float max_lt	(	v16float	v1,
		v16float	v2,
		unsigned int &	cmp
	)

Vector comparison. Returns max(v1,v2) along with lane by lane comparison v1.lane0 < v2.lane0.

Parameters

v1	Vector a
v2	Vector b
cmp	Result of operation a<b

Returns: Result of operation max(a,b)

◆ max_lt() [2/14]

v16int32 max_lt	(	v16int32	a,
		v16int32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [3/14]

v16int32 max_lt	(	v16int32	a,
		v16int32	b,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [4/14]

v16uint32 max_lt	(	v16uint32	a,
		v16uint32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ max_lt() [5/14]

v16uint32 max_lt	(	v16uint32	a,
		v16uint32	b,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [6/14]

v32bfloat16 max_lt	(	v32bfloat16	a,
		v32bfloat16	b,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [7/14]

v32int16 max_lt	(	v32int16	a,
		v32int16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [8/14]

v32int16 max_lt	(	v32int16	a,
		v32int16	b,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [9/14]

v32uint16 max_lt	(	v32uint16	a,
		v32uint16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ max_lt() [10/14]

v32uint16 max_lt	(	v32uint16	a,
		v32uint16	b,
		unsigned int &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [11/14]

v64int8 max_lt	(	v64int8	a,
		v64int8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [12/14]

v64int8 max_lt	(	v64int8	a,
		v64int8	b,
		unsigned long long &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ max_lt() [13/14]

v64uint8 max_lt	(	v64uint8	a,
		v64uint8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ max_lt() [14/14]

v64uint8 max_lt	(	v64uint8	a,
		v64uint8	b,
		unsigned long long &	cmp
	)

Calculates the maximum between two input vectors. Also checks for which vector lanes (a < b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a, b)

◆ maxdiff() [1/12]

v16int32 maxdiff	(	v16int32	a,
		v16int32	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [2/12]

v16int32 maxdiff	(	v16int32	a,
		v16int32	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [3/12]

v16uint32 maxdiff	(	v16uint32	a,
		v16uint32	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [4/12]

v16uint32 maxdiff	(	v16uint32	a,
		v16uint32	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [5/12]

v32int16 maxdiff	(	v32int16	a,
		v32int16	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [6/12]

v32int16 maxdiff	(	v32int16	a,
		v32int16	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [7/12]

v32uint16 maxdiff	(	v32uint16	a,
		v32uint16	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [8/12]

v32uint16 maxdiff	(	v32uint16	a,
		v32uint16	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [9/12]

v64int8 maxdiff	(	v64int8	a,
		v64int8	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [10/12]

v64int8 maxdiff	(	v64int8	a,
		v64int8	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [11/12]

v64uint8 maxdiff	(	v64uint8	a,
		v64uint8	b
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff() [12/12]

v64uint8 maxdiff	(	v64uint8	a,
		v64uint8	b,
		bool	sgn
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [1/12]

v16int32 maxdiff_lt	(	v16int32	a,
		v16int32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [2/12]

v16int32 maxdiff_lt	(	v16int32	a,
		v16int32	b,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [3/12]

v16uint32 maxdiff_lt	(	v16uint32	a,
		v16uint32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ maxdiff_lt() [4/12]

v16uint32 maxdiff_lt	(	v16uint32	a,
		v16uint32	b,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [5/12]

v32int16 maxdiff_lt	(	v32int16	a,
		v32int16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [6/12]

v32int16 maxdiff_lt	(	v32int16	a,
		v32int16	b,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [7/12]

v32uint16 maxdiff_lt	(	v32uint16	a,
		v32uint16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ maxdiff_lt() [8/12]

v32uint16 maxdiff_lt	(	v32uint16	a,
		v32uint16	b,
		unsigned int &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [9/12]

v64int8 maxdiff_lt	(	v64int8	a,
		v64int8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [10/12]

v64int8 maxdiff_lt	(	v64int8	a,
		v64int8	b,
		unsigned long long &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ maxdiff_lt() [11/12]

v64uint8 maxdiff_lt	(	v64uint8	a,
		v64uint8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ maxdiff_lt() [12/12]

v64uint8 maxdiff_lt	(	v64uint8	a,
		v64uint8	b,
		unsigned long long &	cmp
	)

Calculates the difference between two input vectors and returns the maximum between the difference and zero. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of operation: max(a-b, 0)

◆ min() [1/14]

v16float min	(	v16float	v1,
		v16float	v2
	)

Vector comparison. Returns min(v1,v2)

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation min(a,b)

◆ min() [2/14]

v16int32 min	(	v16int32	a,
		v16int32	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [3/14]

v16int32 min	(	v16int32	a,
		v16int32	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min() [4/14]

v16uint32 min	(	v16uint32	a,
		v16uint32	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [5/14]

v16uint32 min	(	v16uint32	a,
		v16uint32	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min() [6/14]

v32bfloat16 min	(	v32bfloat16	a,
		v32bfloat16	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [7/14]

v32int16 min	(	v32int16	a,
		v32int16	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [8/14]

v32int16 min	(	v32int16	a,
		v32int16	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min() [9/14]

v32uint16 min	(	v32uint16	a,
		v32uint16	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [10/14]

v32uint16 min	(	v32uint16	a,
		v32uint16	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min() [11/14]

v64int8 min	(	v64int8	a,
		v64int8	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [12/14]

v64int8 min	(	v64int8	a,
		v64int8	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min() [13/14]

v64uint8 min	(	v64uint8	a,
		v64uint8	b
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b

Returns: Result of operation: min(a, b)

◆ min() [14/14]

v64uint8 min	(	v64uint8	a,
		v64uint8	b,
		bool	sgn
	)

Calculates the minimum between two input vectors.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned

Returns: Result of operation: min(a, b)

◆ min_abs()

v16float min_abs	(	v16float	v1,
		v16float	v2
	)

Vector comparison.

Parameters

v1	Vector a
v2	Vector b

Returns: Result of operation min(a,abs(b))

◆ min_ge() [1/14]

v16float min_ge	(	v16float	v1,
		v16float	v2,
		unsigned int &	cmp
	)

Vector comparison. Returns min(v1,v2) along with lane by lane comparison v1.lane >= v2.lane.

Parameters

v1	Vector a
v2	Vector b
cmp	Result of operation a>=b

Returns: Result of operation min(a,b)

◆ min_ge() [2/14]

v16int32 min_ge	(	v16int32	a,
		v16int32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [3/14]

v16int32 min_ge	(	v16int32	a,
		v16int32	b,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [4/14]

v16uint32 min_ge	(	v16uint32	a,
		v16uint32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ min_ge() [5/14]

v16uint32 min_ge	(	v16uint32	a,
		v16uint32	b,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [6/14]

v32bfloat16 min_ge	(	v32bfloat16	a,
		v32bfloat16	b,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [7/14]

v32int16 min_ge	(	v32int16	a,
		v32int16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [8/14]

v32int16 min_ge	(	v32int16	a,
		v32int16	b,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [9/14]

v32uint16 min_ge	(	v32uint16	a,
		v32uint16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ min_ge() [10/14]

v32uint16 min_ge	(	v32uint16	a,
		v32uint16	b,
		unsigned int &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [11/14]

v64int8 min_ge	(	v64int8	a,
		v64int8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [12/14]

v64int8 min_ge	(	v64int8	a,
		v64int8	b,
		unsigned long long &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ min_ge() [13/14]

v64uint8 min_ge	(	v64uint8	a,
		v64uint8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ min_ge() [14/14]

v64uint8 min_ge	(	v64uint8	a,
		v64uint8	b,
		unsigned long long &	cmp
	)

Calculates the minimum between two input vectors. Also checks for which vector lanes (a >= b) is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of operation: min(a, b)

◆ ne()

unsigned int ne	(	v16float	a,
		v16float	b
	)

Checks for which vector lanes a != b is fulfilled.

Parameters

a	Vector a
b	Vector b

Returns: Result of: a != b

◆ operator!=()

unsigned int operator!=	(	v16float	a,
		v16float	b
	)

Parameters

a	Vector a
b	Vector b

Returns: Result of: a != b

◆ operator<()

unsigned int operator<	(	v16float	a,
		v16float	b
	)

Parameters

a	Vector a
b	Vector b

Returns: Result of: a < b

◆ operator==()

unsigned int operator==	(	v16float	a,
		v16float	b
	)

Parameters

a	Vector a
b	Vector b

Returns: Result of: a == b

◆ operator>()

unsigned int operator>	(	v16float	a,
		v16float	b
	)

Parameters

a	Vector a
b	Vector b

Returns: Result of: a > b

◆ operator>=()

unsigned int operator>=	(	v16float	a,
		v16float	b
	)

Parameters

a	Vector a
b	Vector b

Returns: Result of: a >= b

◆ sel()

v16float sel	(	v16float	a,
		v16float	b,
		unsigned int	sel
	)

Performs lane wise selection between vectors a and b. If a bit of sel if zero the lane of vector a is selected, else the lane of vector b is selected.

Parameters

a	Vector a
b	Vector b
sel	Selection condition.

Returns: Result of the selection

◆ sub_ge() [1/12]

v16int32 sub_ge	(	v16int32	a,
		v16int32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [2/12]

v16int32 sub_ge	(	v16int32	a,
		v16int32	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [3/12]

v16uint32 sub_ge	(	v16uint32	a,
		v16uint32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [4/12]

v16uint32 sub_ge	(	v16uint32	a,
		v16uint32	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [5/12]

v32int16 sub_ge	(	v32int16	a,
		v32int16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [6/12]

v32int16 sub_ge	(	v32int16	a,
		v32int16	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [7/12]

v32uint16 sub_ge	(	v32uint16	a,
		v32uint16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [8/12]

v32uint16 sub_ge	(	v32uint16	a,
		v32uint16	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [9/12]

v64int8 sub_ge	(	v64int8	a,
		v64int8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [10/12]

v64int8 sub_ge	(	v64int8	a,
		v64int8	b,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [11/12]

v64uint8 sub_ge	(	v64uint8	a,
		v64uint8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_ge() [12/12]

v64uint8 sub_ge	(	v64uint8	a,
		v64uint8	b,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a >= b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a >= b

Returns: Result of subtraction (a-b)

◆ sub_lt() [1/12]

v16int32 sub_lt	(	v16int32	a,
		v16int32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [2/12]

v16int32 sub_lt	(	v16int32	a,
		v16int32	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [3/12]

v16uint32 sub_lt	(	v16uint32	a,
		v16uint32	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [4/12]

v16uint32 sub_lt	(	v16uint32	a,
		v16uint32	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [5/12]

v32int16 sub_lt	(	v32int16	a,
		v32int16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [6/12]

v32int16 sub_lt	(	v32int16	a,
		v32int16	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [7/12]

v32uint16 sub_lt	(	v32uint16	a,
		v32uint16	b,
		bool	sgn,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [8/12]

v32uint16 sub_lt	(	v32uint16	a,
		v32uint16	b,
		unsigned int &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [9/12]

v64int8 sub_lt	(	v64int8	a,
		v64int8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [10/12]

v64int8 sub_lt	(	v64int8	a,
		v64int8	b,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [11/12]

v64uint8 sub_lt	(	v64uint8	a,
		v64uint8	b,
		bool	sgn,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
sgn	Sign flag. If it is one, vectors a and b are signed else they are unsigned
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

◆ sub_lt() [12/12]

v64uint8 sub_lt	(	v64uint8	a,
		v64uint8	b,
		unsigned long long &	cmp
	)

Performs vector subraction (a-b) and returns the result. Also checks for which vector lanes a < b is fulfilled.

Parameters

a	Vector a
b	Vector b
cmp	Result of: a < b

Returns: Result of subtraction (a-b)

Overview

sub_lt intrinsic example

Functions

Vector min/max and combined intrinsics for datatype v64uint8

Vector min/max and combined intrinsics for datatype v64int8

Vector min/max and combined intrinsics for datatype v32uint16

Vector min/max and combined intrinsics for datatype v32int16

Vector min/max and combined intrinsics for datatype v16uint32

Vector min/max and combined intrinsics for datatype v16int32

Vector min/max and combined intrinsics for datatype v32bfloat16

Vector min/max and combined intrinsics for datatype v16float

Function Documentation

◆ abs()

◆ eq()

◆ ge()

◆ ge_abs()

◆ gt()

◆ le()

◆ lt()

◆ lt_abs()

◆ max() [1/14]

◆ max() [2/14]

◆ max() [3/14]

◆ max() [4/14]

◆ max() [5/14]

◆ max() [6/14]

◆ max() [7/14]

◆ max() [8/14]

◆ max() [9/14]

◆ max() [10/14]

◆ max() [11/14]

◆ max() [12/14]

◆ max() [13/14]

◆ max() [14/14]

◆ max_abs()

◆ max_lt() [1/14]

◆ max_lt() [2/14]

◆ max_lt() [3/14]

◆ max_lt() [4/14]

◆ max_lt() [5/14]

◆ max_lt() [6/14]

◆ max_lt() [7/14]

◆ max_lt() [8/14]

◆ max_lt() [9/14]

◆ max_lt() [10/14]

◆ max_lt() [11/14]

◆ max_lt() [12/14]

◆ max_lt() [13/14]

◆ max_lt() [14/14]

◆ maxdiff() [1/12]

◆ maxdiff() [2/12]

◆ maxdiff() [3/12]

◆ maxdiff() [4/12]

◆ maxdiff() [5/12]

◆ maxdiff() [6/12]

◆ maxdiff() [7/12]

◆ maxdiff() [8/12]

◆ maxdiff() [9/12]

◆ maxdiff() [10/12]

◆ maxdiff() [11/12]

◆ maxdiff() [12/12]

◆ maxdiff_lt() [1/12]

◆ maxdiff_lt() [2/12]

◆ maxdiff_lt() [3/12]

◆ maxdiff_lt() [4/12]

◆ maxdiff_lt() [5/12]

◆ maxdiff_lt() [6/12]

◆ maxdiff_lt() [7/12]

◆ maxdiff_lt() [8/12]

◆ maxdiff_lt() [9/12]

◆ maxdiff_lt() [10/12]

◆ maxdiff_lt() [11/12]

◆ maxdiff_lt() [12/12]

◆ min() [1/14]

◆ min() [2/14]

◆ min() [3/14]

◆ min() [4/14]

◆ min() [5/14]

◆ min() [6/14]

◆ min() [7/14]