* 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 exp.h
*/
#ifndef LIB_MATH_EXP_H
#define LIB_MATH_EXP_H
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 2201 || __NPU_ARCH__ == 2002 || __NPU_ARCH__ == 3101 || __NPU_ARCH__ == 5102)
#include "kernel_tensor.h"
#if defined(__NPU_ARCH__) && (__NPU_ARCH__ == 3101 || __NPU_ARCH__ == 5102)
#include "../../../impl/adv_api/detail/math/exp/exp_c310_impl.h"
#else
#include "../../../impl/adv_api/detail/math/exp/exp_common_impl.h"
#endif
namespace AscendC {
#pragma begin_pipe(V)
* \ingroup Exp
* \brief Use Taylor Formula for Exp calculation
* exp(a) = (e^b) * (e^c) b is integer part of a, c is decimal part of a. a = b + c
* e^c = 1 + c + c^2 / (2!) + .... + c^n / n!
* \tparam T: Data type to be calculated, half or float
* \tparam taylorExpandLevel: The number of Taylor formula terms (n above)
* \tparam isReuseSource: Whether to reuse the buffer of srcTensor.
* If the value is true, srcTensor can used as tmpBuffer and the data in it will be overwritten.
* If the value is false, srcTensor will not be used as tmpBuffer for calculation.
* \param [out] dstLocal: Output localTensor.
* \param [in] srcLocal: Input localTensor
* \param [in] calCount: The number of elements to be processed.
*/
template <typename T, uint8_t taylorExpandLevel, bool isReuseSource = false>
__aicore__ inline void Exp(const LocalTensor<T>& dstLocal, const LocalTensor<T>& srcLocal, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
static_assert((std::is_same<T, float>::value || std::is_same<T, half>::value),
"Failed to check the data types, current api support data types are half/float.");
ExpAPI::ExpImpl<T, taylorExpandLevel, isReuseSource>(dstLocal, srcLocal, calCount);
}
* \ingroup Exp
* \brief Use Taylor Formula for Exp calculation
* exp(a) = (e^b) * (e^c) b is integer part of a, c is decimal part of a. a = b + c
* e^c = 1 + c + c^2 / (2!) + .... + c^n / n!
* \tparam T: Data type to be calculated, half or float
* \tparam taylorExpandLevel: The number of Taylor formula terms (n above)
* \tparam isReuseSource: Whether to reuse the buffer of srcTensor.
* If the value is true, srcTensor can used as tmpBuffer and the data in it will be overwritten.
* If the value is false, srcTensor will not be used as tmpBuffer for calculation.
* \param [out] dstLocal: Output localTensor.
* \param [in] srcLocal: Input 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 exp_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: The number of elements to be processed.
*/
template <typename T, uint8_t taylorExpandLevel, bool isReuseSource = false>
__aicore__ inline void Exp(const LocalTensor<T>& dstLocal, const LocalTensor<T>& srcLocal,
const LocalTensor<uint8_t>& sharedTmpBuffer, const uint32_t calCount)
{
if ASCEND_IS_AIC {
return;
}
static_assert((std::is_same<T, float>::value || std::is_same<T, half>::value),
"Failed to check the data types, current api support data types are half/float.");
ExpAPI::ExpImpl<T, taylorExpandLevel, isReuseSource>(dstLocal, srcLocal, sharedTmpBuffer, calCount);
}
#pragma end_pipe
}
#endif
#endif
