* Copyright (c) 2025 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License").
* Please refer to the License for details. You may not use this file except in compliance with the License.
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
* \file power.h
* \brief Power function takes the power of base with exponent and returns a tensor with the result.
* Mathematical formulas: Power(x, y) = x ^ y
*/
#if !defined(__ASCENDC_INCLUDE_INTERNAL_HEADERS__)
#define __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#define __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_POWER_H__
#endif
#ifndef LIB_MATH_POWER_H
#define LIB_MATH_POWER_H
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2201 || __NPU_ARCH__ == 2002 || __NPU_ARCH__ == 3510 || \
__NPU_ARCH__ == 5102 || __NPU_ARCH__ == 3003 || __NPU_ARCH__ == 3113)
#include "kernel_tensor.h"
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2002 || __NPU_ARCH__ == 2201)
#include "../../../impl/adv_api/detail/math/power/power_common_impl.h"
#elif defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3510 || __NPU_ARCH__ == 5102 || __NPU_ARCH__ == 3003 || \
__NPU_ARCH__ == 3113)
#include "../../../impl/adv_api/detail/math/power/power_common_utils.h"
#include "../../../impl/adv_api/detail/math/power/power_3510_impl.h"
#endif
namespace AscendC {
#pragma begin_pipe(V)
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3510 || __NPU_ARCH__ == 5102 || __NPU_ARCH__ == 3003 || \
__NPU_ARCH__ == 3113)
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Tensor, src1Tensor, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer)
{
Power<T, isReuseSource, config>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const LocalTensor<T>& src1Tensor)
{
Power<T, isReuseSource, config>(dstTensor, src0Tensor, src1Tensor, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar,
const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Tensor, src1Scalar, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Tensor, src1Scalar, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar,
const LocalTensor<uint8_t>& sharedTmpBuffer)
{
Power<T, isReuseSource, config>(dstTensor, src0Tensor, src1Scalar, sharedTmpBuffer, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar)
{
Power<T, isReuseSource, config>(dstTensor, src0Tensor, src1Scalar, src0Tensor.GetSize());;
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor,
const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Scalar, src1Tensor, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: input Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Scalar, src1Tensor, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor,
const LocalTensor<uint8_t>& sharedTmpBuffer)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Scalar, src1Tensor, sharedTmpBuffer, src1Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
*/
template <typename T, bool isReuseSource = false, const PowerConfig& config = defaultPowerConfig>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor)
{
PowerCommonImpl<T, isReuseSource, config>(dstTensor, src0Scalar, src1Tensor, src1Tensor.GetSize());
}
#else
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor,
const LocalTensor<T>& src1Tensor, const LocalTensor<uint8_t>& sharedTmpBuffer)
{
Power<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, sharedTmpBuffer, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Tensor: exponent LocalTensor
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const LocalTensor<T>& src1Tensor)
{
Power<T, isReuseSource>(dstTensor, src0Tensor, src1Tensor, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar,
const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Tensor, src1Scalar, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Tensor, src1Scalar, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar,
const LocalTensor<uint8_t>& sharedTmpBuffer)
{
Power<T, isReuseSource>(dstTensor, src0Tensor, src1Scalar, sharedTmpBuffer, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Tensor: base LocalTensor
* \param [in] src1Scalar: exponent Scalar
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const LocalTensor<T>& src0Tensor, const T& src1Scalar)
{
Power<T, isReuseSource>(dstTensor, src0Tensor, src1Scalar, src0Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor,
const LocalTensor<uint8_t>& sharedTmpBuffer, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Scalar, src1Tensor, sharedTmpBuffer, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: input Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] calCount: amount of output data to be calculated
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(
const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor, uint32_t calCount)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Scalar, src1Tensor, calCount);
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
* \param [in] sharedTmpBuffer：extra temporary shared space used for intermediate values among calculation process,
* whose required space size should refer to corresponding tiling API, which is defined at
* power_tiling.h. Generally, the more space you allocate, the better performance you will achieve,
* and the performance reaches peak when buffer size is maximum(calculated by tiling function). Moreover, it
* is not guaranteed that the shared space will be cleared after usage, the data could be anything.
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor,
const LocalTensor<uint8_t>& sharedTmpBuffer)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Scalar, src1Tensor, sharedTmpBuffer, src1Tensor.GetSize());
}
* \ingroup Power
* \brief compute Power elementwisely.
* \tparam T: input dataType, support half/float
* \tparam isReuseSource: whether allows API to modify source data, usually for performance reason,
* not enabled currently.
* \param [out] dstTensor: output LocalTensor
* \param [in] src0Scalar: base Scalar
* \param [in] src1Tensor: exponent LocalTensor
*/
template <typename T, bool isReuseSource = false>
__aicore__ inline void Power(const LocalTensor<T>& dstTensor, const T& src0Scalar, const LocalTensor<T>& src1Tensor)
{
PowerCommonImpl<T, isReuseSource>(dstTensor, src0Scalar, src1Tensor, src1Tensor.GetSize());
}
#endif
#pragma end_pipe
}
#endif
#endif
#if defined(__UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_POWER_H__)
#undef __ASCENDC_INCLUDE_INTERNAL_HEADERS__
#undef __UNDEF_ASCENDC_INCLUDE_INTERNAL_HEADERS_POWER_H__
#endif
