| Package | Description |
|---|---|
| org.bytedeco.cuda.cudart | |
| org.bytedeco.cuda.global |
| Modifier and Type | Method and Description |
|---|---|
__nv_bfloat16 |
__nv_bfloat16.put(__nv_bfloat16_raw hr)
\ingroup CUDA_MATH__BFLOAT16_MISC
Assignment operator from \p __nv_bfloat16_raw.
|
__nv_bfloat16 |
__nv_bfloat16.put(double f)
\ingroup CUDA_MATH__BFLOAT16_MISC
Type cast to \p __nv_bfloat16 assignment operator from \p double input using default round-to-nearest-even rounding mode.
|
__nv_bfloat16 |
__nv_bfloat16.put(float f)
\ingroup CUDA_MATH__BFLOAT16_MISC
Type cast to \p __nv_bfloat16 assignment operator from \p float input using default round-to-nearest-even rounding mode.
|
__nv_bfloat16 |
__nv_bfloat16.put(int val)
\ingroup CUDA_MATH__BFLOAT16_MISC
Type cast from \p int assignment operator, using default round-to-nearest-even rounding mode.
|
__nv_bfloat16 |
__nv_bfloat16.put(long val)
\ingroup CUDA_MATH__BFLOAT16_MISC
Type cast from \p long \p long assignment operator, using default round-to-nearest-even rounding mode.
|
__nv_bfloat16 |
__nv_bfloat16.put(short val)
\ingroup CUDA_MATH__BFLOAT16_MISC
Type cast from \p short assignment operator, using default round-to-nearest-even rounding mode.
|
__nv_bfloat16 |
__nv_bfloat162.x()
Storage field holding lower \p __nv_bfloat16 part.
|
__nv_bfloat16 |
__nv_bfloat162.y()
Storage field holding upper \p __nv_bfloat16 part.
|
| Modifier and Type | Method and Description |
|---|---|
__nv_bfloat162 |
__nv_bfloat162.x(__nv_bfloat16 setter) |
__nv_bfloat162 |
__nv_bfloat162.y(__nv_bfloat16 setter) |
| Constructor and Description |
|---|
__half(__nv_bfloat16 f)
\ingroup CUDA_MATH__HALF_MISC
Construct \p __half from \p __nv_bfloat16 input using default round-to-nearest-even rounding mode.
|
__nv_bfloat162(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_MISC
Constructor from two \p __nv_bfloat16 variables
|
| Modifier and Type | Method and Description |
|---|---|
static __nv_bfloat16 |
cudart.__double2bfloat16(double a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts double number to nv_bfloat16 precision in round-to-nearest-even mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__float2bfloat16_rd(float a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts float number to nv_bfloat16 precision in round-down mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__float2bfloat16_rn(float a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts float number to nv_bfloat16 precision in round-to-nearest-even mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__float2bfloat16_ru(float a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts float number to nv_bfloat16 precision in round-up mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__float2bfloat16_rz(float a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts float number to nv_bfloat16 precision in round-towards-zero mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__float2bfloat16(float a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts float number to nv_bfloat16 precision in round-to-nearest-even mode
and returns \p nv_bfloat16 with converted value.
|
static __nv_bfloat16 |
cudart.__habs(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Calculates the absolute value of input \p nv_bfloat16 number and returns the result.
|
static __nv_bfloat16 |
cudart.__hadd_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hadd_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hadd(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hdiv(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 division in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__high2bfloat16(__nv_bfloat162 a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Returns high 16 bits of \p nv_bfloat162 input.
|
static __nv_bfloat16 |
cudart.__hmax_nan(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 maximum of two input values, NaNs pass through.
|
static __nv_bfloat16 |
cudart.__hmax(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 maximum of two input values.
|
static __nv_bfloat16 |
cudart.__hmin_nan(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 minimum of two input values, NaNs pass through.
|
static __nv_bfloat16 |
cudart.__hmin(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 minimum of two input values.
|
static __nv_bfloat16 |
cudart.__hmul_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hmul_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hmul(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hneg(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Negates input \p nv_bfloat16 number and returns the result.
|
static __nv_bfloat16 |
cudart.__hsub_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hsub_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hsub(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__int2bfloat16_rn(int i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a signed integer to a nv_bfloat16 in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__ll2bfloat16_rn(long i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a signed 64-bit integer to a nv_bfloat16 in round-to-nearest-even
mode.
|
static __nv_bfloat16 |
cudart.__low2bfloat16(__nv_bfloat162 a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Returns low 16 bits of \p nv_bfloat162 input.
|
static __nv_bfloat16 |
cudart.__short_as_bfloat16(short i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Reinterprets bits in a signed short integer as a \p nv_bfloat16.
|
static __nv_bfloat16 |
cudart.__short2bfloat16_rn(short i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a signed short integer to a nv_bfloat16 in round-to-nearest-even
mode.
|
static __nv_bfloat16 |
cudart.__uint2bfloat16_rn(int i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert an unsigned integer to a nv_bfloat16 in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__ull2bfloat16_rn(long i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert an unsigned 64-bit integer to a nv_bfloat16 in round-to-nearest-even
mode.
|
static __nv_bfloat16 |
cudart.__ushort_as_bfloat16(short i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Reinterprets bits in an unsigned short integer as a \p nv_bfloat16.
|
static __nv_bfloat16 |
cudart.__ushort2bfloat16_rn(short i)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert an unsigned short integer to a nv_bfloat16 in round-to-nearest-even
mode.
|
static __nv_bfloat16 |
cudart.add(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 addition operation.
|
static __nv_bfloat16 |
cudart.add(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Implements \p nv_bfloat16 unary plus operator, returns input value.
|
static __nv_bfloat16 |
cudart.addPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with addition operation.
|
static __nv_bfloat16 |
cudart.decrement(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 prefix decrement operation.
|
static __nv_bfloat16 |
cudart.decrement(__nv_bfloat16 h,
int ignored)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 postfix decrement operation.
|
static __nv_bfloat16 |
cudart.divide(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 division operation.
|
static __nv_bfloat16 |
cudart.dividePut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with division operation.
|
static __nv_bfloat16 |
cudart.increment(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 prefix increment operation.
|
static __nv_bfloat16 |
cudart.increment(__nv_bfloat16 h,
int ignored)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 postfix increment operation.
|
static __nv_bfloat16 |
cudart.multiply(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 multiplication operation.
|
static __nv_bfloat16 |
cudart.multiplyPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with multiplication operation.
|
static __nv_bfloat16 |
cudart.subtract(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 subtraction operation.
|
static __nv_bfloat16 |
cudart.subtract(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Implements \p nv_bfloat16 unary minus operator.
|
static __nv_bfloat16 |
cudart.subtractPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with subtraction operation.
|
| Modifier and Type | Method and Description |
|---|---|
static short |
cudart.__bfloat16_as_short(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Reinterprets bits in a \p nv_bfloat16 as a signed short integer.
|
static short |
cudart.__bfloat16_as_ushort(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Reinterprets bits in a \p nv_bfloat16 as an unsigned short integer.
|
static __nv_bfloat162 |
cudart.__bfloat162bfloat162(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Returns \p nv_bfloat162 with both halves equal to the input value.
|
static byte |
cudart.__bfloat162char_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to a signed char in round-towards-zero mode.
|
static float |
cudart.__bfloat162float(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Converts \p nv_bfloat16 number to float.
|
static int |
cudart.__bfloat162int_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to a signed integer in round-towards-zero mode.
|
static long |
cudart.__bfloat162ll_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to a signed 64-bit integer in round-towards-zero mode.
|
static short |
cudart.__bfloat162short_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to a signed short integer in round-towards-zero mode.
|
static byte |
cudart.__bfloat162uchar_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to an unsigned char in round-towards-zero mode.
|
static int |
cudart.__bfloat162uint_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to an unsigned integer in round-towards-zero mode.
|
static long |
cudart.__bfloat162ull_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to an unsigned 64-bit integer in round-towards-zero
mode.
|
static short |
cudart.__bfloat162ushort_rz(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Convert a nv_bfloat16 to an unsigned short integer in round-towards-zero
mode.
|
static __nv_bfloat16 |
cudart.__habs(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Calculates the absolute value of input \p nv_bfloat16 number and returns the result.
|
static __nv_bfloat16 |
cudart.__hadd_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hadd_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hadd(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 addition in round-to-nearest-even mode.
|
static __nv_bfloat162 |
cudart.__halves2bfloat162(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Combines two \p nv_bfloat16 numbers into one \p nv_bfloat162 number.
|
static __nv_bfloat16 |
cudart.__hdiv(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 division in round-to-nearest-even mode.
|
static boolean |
cudart.__heq(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 if-equal comparison.
|
static boolean |
cudart.__hequ(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered if-equal comparison.
|
static boolean |
cudart.__hge(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 greater-equal comparison.
|
static boolean |
cudart.__hgeu(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered greater-equal comparison.
|
static boolean |
cudart.__hgt(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 greater-than comparison.
|
static boolean |
cudart.__hgtu(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered greater-than comparison.
|
static int |
cudart.__hisinf(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Checks if the input \p nv_bfloat16 number is infinite.
|
static boolean |
cudart.__hisnan(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Determine whether \p nv_bfloat16 argument is a NaN.
|
static boolean |
cudart.__hle(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 less-equal comparison.
|
static boolean |
cudart.__hleu(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered less-equal comparison.
|
static boolean |
cudart.__hlt(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 less-than comparison.
|
static boolean |
cudart.__hltu(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered less-than comparison.
|
static __nv_bfloat16 |
cudart.__hmax_nan(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 maximum of two input values, NaNs pass through.
|
static __nv_bfloat16 |
cudart.__hmax(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 maximum of two input values.
|
static __nv_bfloat16 |
cudart.__hmin_nan(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 minimum of two input values, NaNs pass through.
|
static __nv_bfloat16 |
cudart.__hmin(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Calculates \p nv_bfloat16 minimum of two input values.
|
static __nv_bfloat16 |
cudart.__hmul_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hmul_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hmul(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 multiplication in round-to-nearest-even mode.
|
static boolean |
cudart.__hne(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 not-equal comparison.
|
static __nv_bfloat16 |
cudart.__hneg(__nv_bfloat16 a)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Negates input \p nv_bfloat16 number and returns the result.
|
static boolean |
cudart.__hneu(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
\brief Performs \p nv_bfloat16 unordered not-equal comparison.
|
static __nv_bfloat16 |
cudart.__hsub_rn(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode.
|
static __nv_bfloat16 |
cudart.__hsub_sat(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode, with
saturation to [0.0, 1.0].
|
static __nv_bfloat16 |
cudart.__hsub(__nv_bfloat16 a,
__nv_bfloat16 b)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
\brief Performs \p nv_bfloat16 subtraction in round-to-nearest-even mode.
|
static int |
cudart.__internal_bfloat162int_rz(__nv_bfloat16 h) |
static int |
cudart.__internal_bfloat162uint_rz(__nv_bfloat16 h) |
static __nv_bfloat16 |
cudart.add(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 addition operation.
|
static __nv_bfloat16 |
cudart.add(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Implements \p nv_bfloat16 unary plus operator, returns input value.
|
static __nv_bfloat16 |
cudart.addPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with addition operation.
|
static int |
cublas.cublasTSSgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
float[] alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
float[] beta,
float[] yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSSgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatBuffer alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
FloatBuffer beta,
FloatBuffer yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSSgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatPointer alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
FloatPointer beta,
FloatPointer yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSSgemvBatched(cublasContext handle,
int trans,
int m,
int n,
float[] alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
float[] beta,
float[] yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSSgemvBatched(cublasContext handle,
int trans,
int m,
int n,
FloatBuffer alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
FloatBuffer beta,
FloatBuffer yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSSgemvBatched(cublasContext handle,
int trans,
int m,
int n,
FloatPointer alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
FloatPointer beta,
FloatPointer yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
float[] alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
float[] beta,
float[] y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatBuffer alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
FloatBuffer beta,
FloatBuffer y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatPointer alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
FloatPointer beta,
FloatPointer y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
float[] alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
float[] beta,
float[] y,
int incy,
long stridey,
int batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
FloatBuffer alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
FloatBuffer beta,
FloatBuffer y,
int incy,
long stridey,
int batchCount) |
static int |
cublas.cublasTSSgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
FloatPointer alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
FloatPointer beta,
FloatPointer y,
int incy,
long stridey,
int batchCount) |
static int |
cublas.cublasTSTgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
float[] alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
float[] beta,
__nv_bfloat16 yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSTgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatBuffer alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
FloatBuffer beta,
__nv_bfloat16 yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSTgemvBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatPointer alpha,
__nv_bfloat16 Aarray,
long lda,
__nv_bfloat16 xarray,
long incx,
FloatPointer beta,
__nv_bfloat16 yarray,
long incy,
long batchCount) |
static int |
cublas.cublasTSTgemvBatched(cublasContext handle,
int trans,
int m,
int n,
float[] alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
float[] beta,
__nv_bfloat16 yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSTgemvBatched(cublasContext handle,
int trans,
int m,
int n,
FloatBuffer alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
FloatBuffer beta,
__nv_bfloat16 yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSTgemvBatched(cublasContext handle,
int trans,
int m,
int n,
FloatPointer alpha,
__nv_bfloat16 Aarray,
int lda,
__nv_bfloat16 xarray,
int incx,
FloatPointer beta,
__nv_bfloat16 yarray,
int incy,
int batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
float[] alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
float[] beta,
__nv_bfloat16 y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatBuffer alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
FloatBuffer beta,
__nv_bfloat16 y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched_64(cublasContext handle,
int trans,
long m,
long n,
FloatPointer alpha,
__nv_bfloat16 A,
long lda,
long strideA,
__nv_bfloat16 x,
long incx,
long stridex,
FloatPointer beta,
__nv_bfloat16 y,
long incy,
long stridey,
long batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
float[] alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
float[] beta,
__nv_bfloat16 y,
int incy,
long stridey,
int batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
FloatBuffer alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
FloatBuffer beta,
__nv_bfloat16 y,
int incy,
long stridey,
int batchCount) |
static int |
cublas.cublasTSTgemvStridedBatched(cublasContext handle,
int trans,
int m,
int n,
FloatPointer alpha,
__nv_bfloat16 A,
int lda,
long strideA,
__nv_bfloat16 x,
int incx,
long stridex,
FloatPointer beta,
__nv_bfloat16 y,
int incy,
long stridey,
int batchCount) |
static __nv_bfloat16 |
cudart.decrement(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 prefix decrement operation.
|
static __nv_bfloat16 |
cudart.decrement(__nv_bfloat16 h,
int ignored)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 postfix decrement operation.
|
static __nv_bfloat16 |
cudart.divide(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 division operation.
|
static __nv_bfloat16 |
cudart.dividePut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with division operation.
|
static boolean |
cudart.equals(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 ordered compare equal operation.
|
static boolean |
cudart.greaterThan(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 ordered greater-than compare operation.
|
static boolean |
cudart.greaterThanEquals(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 ordered greater-or-equal compare operation.
|
static __nv_bfloat16 |
cudart.increment(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 prefix increment operation.
|
static __nv_bfloat16 |
cudart.increment(__nv_bfloat16 h,
int ignored)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 postfix increment operation.
|
static boolean |
cudart.lessThan(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 ordered less-than compare operation.
|
static boolean |
cudart.lessThanEquals(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 ordered less-or-equal compare operation.
|
static __nv_bfloat162 |
cudart.make_bfloat162(__nv_bfloat16 x,
__nv_bfloat16 y)
\ingroup CUDA_MATH__BFLOAT16_MISC
\brief Vector function, combines two \p nv_bfloat16 numbers into one \p nv_bfloat162 number.
|
static __nv_bfloat16 |
cudart.multiply(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 multiplication operation.
|
static __nv_bfloat16 |
cudart.multiplyPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with multiplication operation.
|
static boolean |
cudart.notEquals(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_COMPARISON
Performs \p nv_bfloat16 unordered compare not-equal operation.
|
static __nv_bfloat16 |
cudart.subtract(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 subtraction operation.
|
static __nv_bfloat16 |
cudart.subtract(__nv_bfloat16 h)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Implements \p nv_bfloat16 unary minus operator.
|
static __nv_bfloat16 |
cudart.subtractPut(__nv_bfloat16 lh,
__nv_bfloat16 rh)
\ingroup CUDA_MATH__BFLOAT16_ARITHMETIC
Performs \p nv_bfloat16 compound assignment with subtraction operation.
|
Copyright © 2025. All rights reserved.