CANN电力仿真张量求和优化

aclnnSum【免费下载链接】elec-ops-simulationelec-ops-simulation 是 CANN 社区 Electrical Engineering SIG电力行业兴趣小组旗下的电力仿真求解算子库 聚焦于计算电网在稳态运行条件下各节点的电压、相角以及各支路线路、变压器的功率分布的仿真核心需求面向华为昇腾Ascend硬件平台进行深度优化。项目地址: https://gitcode.com/cann/elec-ops-simulation 查看源码产品支持情况产品是否支持Ascend 950PR/Ascend 950DT×Atlas A3 训练系列产品/Atlas A3 推理系列产品√Atlas A2 训练系列产品/Atlas A2 推理系列产品√Atlas 200I/500 A2 推理产品×Atlas 推理系列产品×Atlas 训练系列产品√功能说明返回输入tensors列表中每个输入tensor依次做add求和。函数原型每个算子分为两段式接口必须先调用“aclnnSumGetWorkspaceSize”接口获取计算所需workspace大小以及包含了算子计算流程的执行器再调用“aclnnSum”接口执行计算。aclnnStatus aclnnSumGetWorkspaceSize( const aclTensorList* tensors, aclTensor* out, uint64_t* workspaceSize, aclOpExecutor** executor)aclnnStatus aclnnSum( void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, aclrtStream stream)aclnnSumGetWorkspaceSize参数说明参数名输入/输出描述使用说明数据类型数据格式维度(shape)非连续TensortensorsaclTensorList*输入需要计算的输入tensors列表。需要与out数据类型相同。tensors中各tensor的shape需要与out满足broadcast关系。FLOAT16、FLOAT、INT8、INT32、UINT8ND不大于8√outaclTensor*输出输出tensor。需要与tensors数据类型相同。shape需要与tensors中各tensor的shape满足broadcast关系。FLOAT、FLOAT16、INT8、INT32、UINT8ND-√workspaceSizeuint64_t*输出返回需要在Device侧申请的workspace大小。-----executoraclOpExecutor**输出返回op执行器包含了算子计算流程。-----返回值aclnnStatus返回状态码具体参见aclnn返回码。第一段接口完成入参校验出现如下场景时报错返回值错误码描述ACLNN_ERR_PARAM_NULLPTR161001传入的tensors或out是空指针时。ACLNN_ERR_PARAM_INVALID161002tensors列表中tensor或out的数据类型不在支持的范围之内。tensors列表和out的数据类型不一致。tensors和out的shape不满足broadcast规则,或者broadcast后的shape与out不一致。aclnnSum参数说明参数名输入/输出描述workspace输入在Device侧申请的workspace内存地址。workspaceSize输入在Device侧申请的workspace大小由第一段接口aclnnSumGetWorkspaceSize获取。executor输入op执行器包含了算子计算流程。stream输入指定执行任务的Stream。返回值aclnnStatus返回状态码具体参见aclnn返回码。约束说明确定性计算aclnnSum默认确定性实现。调用示例示例代码如下仅供参考具体编译和执行过程请参考编译与运行样例。#include iostream #include vector #include acl/acl.h #include aclnnop/aclnn_sum.h #define CHECK_RET(cond, return_expr) \ do { \ if (!(cond)) { \ return_expr; \ } \ } while (0) #define LOG_PRINT(message, ...) \ do { \ printf(message, ##__VA_ARGS__); \ } while (0) int64_t GetShapeSize(const std::vectorint64_t shape) { int64_t shapeSize 1; for (auto i : shape) { shapeSize * i; } return shapeSize; } int Init(int32_t deviceId, aclrtStream* stream) { // 固定写法资源初始化 auto ret aclInit(nullptr); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclInit failed. ERROR: %d\n, ret); return ret); ret aclrtSetDevice(deviceId); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclrtSetDevice failed. ERROR: %d\n, ret); return ret); ret aclrtCreateStream(stream); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclrtCreateStream failed. ERROR: %d\n, ret); return ret); return 0; } template typename T int CreateAclTensor(const std::vectorT hostData, const std::vectorint64_t shape, void** deviceAddr, aclDataType dataType, aclTensor** tensor) { auto size GetShapeSize(shape) * sizeof(T); // 调用aclrtMalloc申请device侧内存 auto ret aclrtMalloc(deviceAddr, size, ACL_MEM_MALLOC_HUGE_FIRST); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclrtMalloc failed. ERROR: %d\n, ret); return ret); // 调用aclrtMemcpy将host侧数据拷贝到device侧内存上 ret aclrtMemcpy(*deviceAddr, size, hostData.data(), size, ACL_MEMCPY_HOST_TO_DEVICE); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclrtMemcpy failed. ERROR: %d\n, ret); return ret); // 计算连续tensor的strides std::vectorint64_t strides(shape.size(), 1); for (int64_t i shape.size() - 2; i 0; i--) { strides[i] shape[i 1] * strides[i 1]; } // 调用aclCreateTensor接口创建aclTensor *tensor aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND, shape.data(), shape.size(), *deviceAddr); return 0; } int main() { // 1. 固定写法device/stream初始化参考acl API手册 // 根据自己的实际device填写deviceId int32_t deviceId 0; aclrtStream stream; auto ret Init(deviceId, stream); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(Init acl failed. ERROR: %d\n, ret); return ret); // 2. 构造输入与输出需要根据API的接口自定义构造 std::vectorint64_t selfShape1 {2, 3}; std::vectorint64_t selfShape2 {2, 3}; std::vectorint64_t outShape {2, 3}; void* input1DeviceAddr nullptr; void* input2DeviceAddr nullptr; void* outDeviceAddr nullptr; aclTensor* input1 nullptr; aclTensor* input2 nullptr; aclTensor* out nullptr; std::vectorfloat input1HostData {1, 2, 3, 4, 5, 6}; std::vectorfloat input2HostData {7, 8, 9, 10, 11, 12}; std::vectorfloat outHostData(6, 0); // 创建input1 aclTensor ret CreateAclTensor(input1HostData, selfShape1, input1DeviceAddr, aclDataType::ACL_FLOAT, input1); CHECK_RET(ret ACL_SUCCESS, return ret); // 创建input2 aclTensor ret CreateAclTensor(input2HostData, selfShape2, input2DeviceAddr, aclDataType::ACL_FLOAT, input2); CHECK_RET(ret ACL_SUCCESS, return ret); // 创建aclTensorList std::vectoraclTensor* tmp{input1, input2}; aclTensorList* tensorList aclCreateTensorList(tmp.data(), tmp.size()); // 创建out aclTensor ret CreateAclTensor(outHostData, outShape, outDeviceAddr, aclDataType::ACL_FLOAT, out); CHECK_RET(ret ACL_SUCCESS, return ret); // 3. 调用CANN算子库API需要修改为具体的API名称 uint64_t workspaceSize 0; aclOpExecutor* executor; // 调用aclnnSum第一段接口 ret aclnnSumGetWorkspaceSize(tensorList, out, workspaceSize, executor); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclnnSumGetWorkspaceSize failed. ERROR: %d\n, ret); return ret); // 根据第一段接口计算出的workspaceSize申请device内存 void* workspaceAddr nullptr; if (workspaceSize 0) { ret aclrtMalloc(workspaceAddr, workspaceSize, ACL_MEM_MALLOC_HUGE_FIRST); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(allocate workspace failed. ERROR: %d\n, ret); return ret); } // 调用aclnnSum第二段接口 ret aclnnSum(workspaceAddr, workspaceSize, executor, stream); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclnnSum failed. ERROR: %d\n, ret); return ret); // 4. 固定写法同步等待任务执行结束 ret aclrtSynchronizeStream(stream); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(aclrtSynchronizeStream failed. ERROR: %d\n, ret); return ret); // 5. 获取输出的值将device侧内存上的结果拷贝至host侧需要根据具体API的接口定义修改 auto size GetShapeSize(outShape); std::vectorfloat outData(size, 0); ret aclrtMemcpy(outData.data(), outData.size() * sizeof(outData[0]), outDeviceAddr, size * sizeof(outData[0]), ACL_MEMCPY_DEVICE_TO_HOST); CHECK_RET(ret ACL_SUCCESS, LOG_PRINT(copy result from device to host failed. ERROR: %d\n, ret); return ret); for (int64_t i 0; i size; i) { LOG_PRINT(out result[%ld] is: %f\n, i, outData[i]); } // 6. 释放aclTensor和aclScalar需要根据具体API的接口定义修改 aclDestroyTensorList(tensorList); aclDestroyTensor(out); // 7. 释放device资源需要根据具体API的接口定义修改 aclrtFree(input1DeviceAddr); aclrtFree(input2DeviceAddr); aclrtFree(outDeviceAddr); if (workspaceSize 0) { aclrtFree(workspaceAddr); } aclrtDestroyStream(stream); aclrtResetDevice(deviceId); aclFinalize(); return 0; }【免费下载链接】elec-ops-simulationelec-ops-simulation 是 CANN 社区 Electrical Engineering SIG电力行业兴趣小组旗下的电力仿真求解算子库 聚焦于计算电网在稳态运行条件下各节点的电压、相角以及各支路线路、变压器的功率分布的仿真核心需求面向华为昇腾Ascend硬件平台进行深度优化。项目地址: https://gitcode.com/cann/elec-ops-simulation创作声明：本文部分内容由AI辅助生成（AIGC），仅供参考