aclnnBatchMatMulWeightNz

Applicable Products

Product	Supported
Ascend 950PR/Ascend 950DT	√
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	×
Atlas inference products	×
Atlas training products	×

Function

Description: Performs matrix multiplication between tensor self and tensor mat2. mat2 only supports the NZ format. Only supports self with 3 dimensions and mat2 with 5 dimensions.
Formula:
$out = self@mat2$

Function Prototype

Each operator uses a two-phase API. You must first call the "aclnnBatchMatMulWeightNzGetWorkspaceSize" API to obtain the input parameters and calculate the required workspace size based on the process, and then call the "aclnnBatchMatMulWeightNz" API to perform the computation.

aclnnStatus aclnnBatchMatMulWeightNzGetWorkspaceSize(
  const aclTensor *self,
  const aclTensor *mat2,
  aclTensor       *out,
  int8_t           cubeMathType,
  uint64_t        *workspaceSize,
  aclOpExecutor   **executor)

aclnnStatus aclnnBatchMatMulWeightNz(
  void            *workspace,
  uint64_t         workspaceSize,
  aclOpExecutor   *executor,
  aclrtStream      stream)

aclnnBatchMatMulWeightNzGetWorkspaceSize

Parameters:

Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Shape	Non-contiguous Tensor
self	Input	Represents the first matrix in matrix multiplication, which is self in the formula.	The data type must satisfy the data type deduction rules with mat2 (see Deduction Relationship and Constraints). When self is not transposed, each dimension represents: (b, m, k). When self is transposed, each dimension represents: (b, k, m). The first dimension b of self must satisfy the broadcast relationship with the first dimension b of mat2.	BFLOAT16, FLOAT16	ND	3	√
mat2	Input	Represents the second matrix in matrix multiplication, i.e., mat2 in the formula.	The data type must satisfy the data type deduction rules with self (see Deduction Relationship and Constraints). The reduce dimension of mat2 must be equal in size to the reduce dimension of self. When matrix B is not transposed, the dimensions in NZ format are: (b, n1, k1, k0, n0), where k0 = 16 and n0 = 16. The k in self shape and k1 in mat2 shape must satisfy: ceil(k, k0) = k1. The n1 in mat2 shape and n in out must satisfy: ceil(n, n0) = n1. When matrix B is transposed, the dimensions in NZ format are: (b, k1, n1, n0, k0), where n0 = 16 and k0 = 16. The k in self shape and k1 in mat2 shape must satisfy: ceil(k, k0) = k1. The n1 in mat2 shape and n in out must satisfy: ceil(n, n0) = n1. The first dimension b of mat2 must satisfy the broadcast relationship with the first dimension b of self.	BFLOAT16, FLOAT16	NZ	5	√
out	Output	Represents the output matrix of the matrix multiplication, i.e., out in the formula.	The data type must be consistent with the data type derived from self and mat2 (see Deduction Relationship and Constraints). Dimensions: (b, m, n), where m is consistent with the m of self, n satisfies ceil(n / n0) = n1 with the n1 and n0 of mat2, and b must be consistent with the b of self and the b of mat2 after broadcast deduction.	BFLOAT16, FLOAT16	ND	3	-
cubeMathType	Input	Specifies the computation logic of the Cube unit.	If the input data types have a deduction relationship, this parameter processes the deduced data type by default. The supported enumeration values are as follows: 0: KEEP_DTYPE. Keeps the input data type for computation. 1: ALLOW_FP32_DOWN_PRECISION. Allows down-precision computation of the input data. This option is not supported. 2: USE_FP16. Allows down-precision of the input to FLOAT16 for computation. This option is not supported when the input data type is BFLOAT16. 3: USE_HF32. Allows down-precision of the input to HFLOAT32 for computation. This option is not supported.	INT8	-	-	-
workspaceSize	Output	Returns the workspace size to be applied for on the Device side.	-	-	-	-	-
executor	Output	Returns the operator executor, which contains the operator computation process.	-	-	-	-	-

Return value

aclnnStatus: return code. For details, see aclnn Return Code.

The first-phase API performs input parameter validation and returns an error in the following scenarios:

Return Value	Error Code	Description
ACLNN_ERR_PARAM_NULLPTR	161001	The input self, mat2, or out is a null pointer.
ACLNN_ERR_PARAM_INVALID	161002	The data types or data formats of self and mat2 are not within the supported range.
		Data type deduction cannot be performed for self and mat2.
		The deduced data type cannot be converted to the specified output type of out.

aclnnBatchMatMulWeightNz

Parameters:

Parameter	Input/Output	Description
workspace	Input	Address of the workspace memory applied for on the Device side.
workspaceSize	Input	Size of the workspace applied for on the Device side, obtained from the first-phase API aclnnBatchMatMulWeightNzGetWorkspaceSize.
executor	Input	Operator executor, which contains the operator computation flow.
stream	Input	Specifies the stream for task execution.

Return value

aclnnStatus: return code. For details, see aclnn Return Code.

Constraints

Deterministic computation:
- aclnnBatchMatMulWeightNz defaults to a deterministic implementation.
Data type deduction between BFLOAT16 and FLOAT16 inputs is not supported.
self only supports 3 dimensions, and mat2 only supports the Ascend private format. Before calling this API, you must complete the conversion of mat2 from ND to the Ascend private format.
The case where one of the last two axes of mat2 is 1, i.e., k=1 or n=1, is not supported.

Example

Atlas A2 training products/Atlas A2 inference products, Atlas A3 training products/Atlas A3 inference products:

The following is sample code for the scenario where self and mat2 have the float16 data type and mat2 is in NZ format. This is for reference only. For details about compilation and execution, see Compile and Run Samples.

#include <iostream>
#include <vector>
#include <cmath>
#include "acl/acl.h"
#include "aclnnop/aclnn_batch_matmul.h"
#include "aclnnop/aclnn_trans_matmul_weight.h"
#include "aclnnop/aclnn_cast.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::vector<int64_t>& shape) {
  int64_t shapeSize = 1;
  for (auto i : shape) {
    shapeSize *= i;
  }
  return shapeSize;
}

// Convert the uint16_t representation of FP16 to a float representation.
float Fp16ToFloat(uint16_t h) {
  int s = (h >> 15) & 0x1;              // sign
  int e = (h >> 10) & 0x1F;             // exponent
  int f =  h        & 0x3FF;            // fraction
  if (e == 0) {
    // Zero or Denormal
    if (f == 0) {
      return s ? -0.0f : 0.0f;
    }
    // Denormals
    float sig = f / 1024.0f;
    float result = sig * pow(2, -24);
    return s ? -result : result;
  } else if (e == 31) {
      // Infinity or NaN
      return f == 0 ? (s ? -INFINITY : INFINITY) : NAN;
  }
  // Normalized
  float result = (1.0f + f / 1024.0f) * pow(2, e - 15);
  return s ? -result : result;
}

int Init(int32_t deviceId, aclrtStream* stream) {
  // Boilerplate for resource initialization.
  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::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                    aclDataType dataType, aclTensor** tensor) {
  auto size = GetShapeSize(shape) * sizeof(T);
  // Call aclrtMalloc to apply for device-side memory.
  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);
  // Copy data from the host to the device memory by calling aclrtMemcpy.
  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);

  // Calculate the strides of a contiguous tensor.
  std::vector<int64_t> strides(shape.size(), 1);
  for (int64_t i = shape.size() - 2; i >= 0; i--) {
    strides[i] = shape[i + 1] * strides[i + 1];
  }

  // Create an aclTensor by calling aclCreateTensor.
  *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND,
                            shape.data(), shape.size(), *deviceAddr);
  return 0;
}

template <typename T>
int CreateAclTensorWeight(const std::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                          aclDataType dataType, aclTensor** tensor) {
  auto size = static_cast<uint64_t>(GetShapeSize(shape));

  const aclIntArray* mat2Size = aclCreateIntArray(shape.data(), shape.size());
  auto ret = aclnnCalculateMatmulWeightSize(mat2Size, &size);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnCalculateMatmulWeightSize failed. ERROR: %d\n", ret); return ret);
  size *= sizeof(T);

  // Allocate device memory by calling aclrtMalloc.
  ret = aclrtMalloc(deviceAddr, size, ACL_MEM_MALLOC_HUGE_FIRST);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMalloc failed. ERROR: %d\n", ret); return ret);
  // Copy data from the host to the device memory by calling aclrtMemcpy.
  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);

  // Calculate the strides of a contiguous tensor.
  std::vector<int64_t> strides(shape.size(), 1);
  for (int64_t i = shape.size() - 2; i >= 0; i--) {
    strides[i] = shape[i + 1] * strides[i + 1];
  }

  std::vector<int64_t> storageShape;
  storageShape.push_back(GetShapeSize(shape));

  // Call the aclCreateTensor API to create an aclTensor.
  *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND,
                            storageShape.data(), storageShape.size(), *deviceAddr);
  return 0;
}

int main() {
  // 1. (Boilerplate) Initialize the device/stream. Refer to the ACL API manual.
  // Enter the actual device ID for 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. Construct the inputs and outputs. Custom construction is required based on the API.
  std::vector<int64_t> selfShape = {2, 16, 32};
  std::vector<int64_t> mat2Shape = {2, 32, 16};
  std::vector<int64_t> outShape = {2, 16, 16};
  void* selfDeviceAddr = nullptr;
  void* mat2DeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  aclTensor* self = nullptr;
  aclTensor* mat2 = nullptr;
  aclTensor* out = nullptr;
  std::vector<uint16_t> selfHostData(1024, 0x3C00); // float16_t uses 0x3C00 to represent 1 in int16.
  std::vector<uint16_t> mat2HostData(1024, 0x3C00); // float16_t uses 0x3C00 to represent 1 in int16.
  std::vector<uint16_t> outHostData(512, 0);
  // Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT16, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the other aclTensor.
  ret = CreateAclTensorWeight(mat2HostData, mat2Shape, &mat2DeviceAddr, aclDataType::ACL_FLOAT16, &mat2);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the out aclTensor.
  ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT16, &out);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. Call the CANN operator library API. Replace with the specific API name.
  int8_t cubeMathType = 0;
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // Call TransWeight.
  ret = aclnnTransMatmulWeightGetWorkspaceSize(mat2, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnTransMatmulWeightGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);

  // Apply for device memory based on the workspaceSize calculated by the first-phase API.
  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);
  }

  // Call the second-phase API of aclnnTransMatmulWeight.
  ret = aclnnTransMatmulWeight(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnTransMatmulWeight failed. ERROR: %d\n", ret); return ret);

  // Call the first-phase API of aclnnBatchMatMulWeightNz.
  uint64_t workspaceSizeMm = 0;
  ret = aclnnBatchMatMulWeightNzGetWorkspaceSize(self, mat2, out, cubeMathType, &workspaceSizeMm, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchMatMulWeightNzGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);

  // Apply for device memory based on the workspaceSize calculated by the first-phase API.
  void* workspaceAddrMm = nullptr;
  if (workspaceSizeMm > 0) {
    ret = aclrtMalloc(&workspaceAddrMm, workspaceSizeMm, ACL_MEM_MALLOC_HUGE_FIRST);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("allocate workspace failed. ERROR: %d\n", ret); return ret);
  }
  // Call the second-phase API of aclnnBatchMatMulWeightNz.
  ret = aclnnBatchMatMulWeightNz(workspaceAddrMm, workspaceSizeMm, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchMatMulWeightNz failed. ERROR: %d\n", ret); return ret);

  // 4. (boilerplate) Synchronize and wait for task execution to complete.
  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  // 5. Obtain the output value and copy the result from the device memory to the host memory. Modify this based on the specific API definition.
  auto size = GetShapeSize(outShape);
  std::vector<uint16_t> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), outDeviceAddr,
                    size * sizeof(resultData[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);
  // In C, fp16 data cannot be printed directly. It must be read as uint16 and manually converted to fp16 in float representation via binary conversion.
  for (int64_t i = 0; i < size; i++) {
    float fp16Float = Fp16ToFloat(resultData[i]);
    LOG_PRINT("result[%ld] is: %f\n", i, fp16Float);
  }

  // 6. Release aclTensor and aclScalar. Modify based on the specific API definition.
  aclDestroyTensor(self);
  aclDestroyTensor(mat2);
  aclDestroyTensor(out);

  // 7. Release device resources. Modify based on the specific API definition.
  aclrtFree(selfDeviceAddr);
  aclrtFree(mat2DeviceAddr);
  aclrtFree(outDeviceAddr);

  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  if (workspaceSizeMm > 0) {
    aclrtFree(workspaceAddrMm);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();
  return 0;
}

Ascend 950PR/Ascend 950DT:

The following shows sample code for the scenario where self and mat2 are of the float16 data type and mat2 is in NZ format. This is for reference only. For details about compilation and execution, see Compile and Run Samples.

#include <iostream>
#include <memory>
#include <vector>
#include <cmath>
#include "acl/acl.h"
#include "aclnnop/aclnn_batch_matmul.h"
#include "aclnnop/aclnn_npu_format_cast.h"
#include "aclnnop/aclnn_cast.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::vector<int64_t>& shape) {
  int64_t shapeSize = 1;
  for (auto i : shape) {
    shapeSize *= i;
  }
  return shapeSize;
}

// Convert the uint16_t representation of FP16 to float representation.
float Fp16ToFloat(uint16_t h) {
  int s = (h >> 15) & 0x1;              // sign
  int e = (h >> 10) & 0x1F;             // exponent
  int f =  h        & 0x3FF;            // fraction
  if (e == 0) {
    // Zero or Denormal
    if (f == 0) {
      return s ? -0.0f : 0.0f;
    }
    // Denormals
    float sig = f / 1024.0f;
    float result = sig * pow(2, -24);
    return s ? -result : result;
  } else if (e == 31) {
      // Infinity or NaN
      return f == 0 ? (s ? -INFINITY : INFINITY) : NAN;
  }
  // Normalized
  float result = (1.0f + f / 1024.0f) * pow(2, e - 15);
  return s ? -result : result;
}

int Init(int32_t deviceId, aclrtStream* stream) {
  // Boilerplate: resource initialization.
  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::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                    aclDataType dataType, aclTensor** tensor) {
  auto size = GetShapeSize(shape) * sizeof(T);
  // Call aclrtMalloc to allocate memory on the device side.
  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);
  // Call aclrtMemcpy to copy data from the host side to the device memory.
  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);

  // Calculate the strides of a contiguous tensor.
  std::vector<int64_t> strides(shape.size(), 1);
  for (int64_t i = shape.size() - 2; i >= 0; i--) {
    strides[i] = shape[i + 1] * strides[i + 1];
  }

  // Call aclCreateTensor to create an aclTensor.
  *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND,
                            shape.data(), shape.size(), *deviceAddr);
  return 0;
}

template <typename T>
int CreateAclTensorWithFormat(
    const std::vector<T>& hostData, const std::vector<int64_t>& shape, int64_t** storageShape,
    uint64_t* storageShapeSize, void** deviceAddr, aclDataType dataType, aclTensor** tensor, aclFormat format)
{
    auto size = hostData.size() * sizeof(T);
    // Call aclrtMalloc to allocate device-side memory.
    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);
    // Call aclrtMemcpy to copy host-side data to device-side memory.
    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);

    // Calculate the strides of a contiguous tensor.
    std::vector<int64_t> strides(shape.size(), 1);
    for (int64_t i = shape.size() - 2; i >= 0; i--) {
        strides[i] = shape[i + 1] * strides[i + 1];
    }

    *tensor = aclCreateTensor(
        shape.data(), shape.size(), dataType, strides.data(), 0, format, *storageShape, *storageShapeSize, *deviceAddr);
    return 0;
}

int main() {
  // 1. (Boilerplate) Initialize device/stream. See the ACL API manual.
  // Enter the deviceId based on the actual device.
  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. Construct the input and output. Customize the construction based on the API interface.
  std::vector<int64_t> selfShape = {2, 16, 32};
  std::vector<int64_t> mat2Shape = {2, 32, 16};
  std::vector<int64_t> outShape = {2, 16, 16};

  std::vector<uint16_t> selfHostData(1024, 0x3C00); // 0x3C00 represents 1 in int16 for float16_t.
  std::vector<uint16_t> mat2HostData(1024, 0x3C00); // 0x3C00 represents 1 in int16 for float16_t.
  std::vector<uint16_t> outHostData(512, 0);

  aclTensor* self = nullptr;
  aclTensor* mat2 = nullptr;
  aclTensor* out = nullptr;
  aclTensor* mat2NZ = nullptr;

  void* selfDeviceAddr = nullptr;
  void* mat2DeviceAddr = nullptr;
  void* outDeviceAddr = nullptr;
  void* dstDeviceAddr = nullptr;

  // Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT16, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the other aclTensor.
  ret = CreateAclTensor(mat2HostData, mat2Shape, &mat2DeviceAddr, aclDataType::ACL_FLOAT16, &mat2);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the out aclTensor.
  ret = CreateAclTensor(outHostData, outShape, &outDeviceAddr, aclDataType::ACL_FLOAT16, &out);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. Convert the weight tensor from ND to NZ by calling the npu_format_cast API.
  aclDataType additionalDtype = aclDataType::ACL_FLOAT16;
  aclDataType srcDtype = aclDataType::ACL_FLOAT16;
  int64_t* dstShape = nullptr;
  uint64_t dstShapeSize = 0;
  int actualFormat;
  int8_t cubeMathType = 0;
  aclOpExecutor* executor = nullptr;

  uint64_t workspaceSize = 0;
  uint64_t workspaceSizeMm = 0;

  void* workspaceAddr = nullptr;
  void* workspaceAddrMm = nullptr;
  // Calculate the shape and format of the target tensor.
  ret = aclnnNpuFormatCastCalculateSizeAndFormat(mat2, 29, additionalDtype, &dstShape, &dstShapeSize, &actualFormat);
  CHECK_RET(
      ret == ACL_SUCCESS, LOG_PRINT("aclnnNpuFormatCastCalculateSizeAndFormat failed. ERROR: %d\n", ret); return ret);

  ret = CreateAclTensorWithFormat(
      mat2HostData, mat2Shape, &dstShape, &dstShapeSize, &dstDeviceAddr, srcDtype, &mat2NZ,
      static_cast<aclFormat>(actualFormat));
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("CreateAclTensorWithFormat failed. ERROR: %d\n", ret); return ret);

  // Call the first-phase API of aclnnNpuFormatCastGetWorkspaceSize.
  ret = aclnnNpuFormatCastGetWorkspaceSize(mat2, mat2NZ, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnNpuFormatCastGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);

  // Allocate device memory based on the workspaceSize calculated by the first-phase API.
  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);
  }

  // Call the second-phase API of aclnnNpuFormatCastGetWorkspaceSize.
  ret = aclnnNpuFormatCast(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnNpuFormatCast failed. ERROR: %d\n", ret); return ret);

  // 4. Synchronize and wait for task execution to complete.
  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  // Call the first-phase API of aclnnBatchMatMulWeightNz.
  ret = aclnnBatchMatMulWeightNzGetWorkspaceSize(self, mat2NZ, out, cubeMathType, &workspaceSizeMm, &executor);
  CHECK_RET(
      ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchMatMulWeightNzGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);

  // Apply for device memory based on the workspaceSize calculated by the first-phase API.
  if (workspaceSizeMm > 0) {
      ret = aclrtMalloc(&workspaceAddrMm, workspaceSizeMm, ACL_MEM_MALLOC_HUGE_FIRST);
      CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("allocate workspace failed. ERROR: %d\n", ret); return ret);
  }
  // Call the second-phase API of aclnnBatchMatMulWeightNz.
  ret = aclnnBatchMatMulWeightNz(workspaceAddrMm, workspaceSizeMm, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnBatchMatMulWeightNz failed. ERROR: %d\n", ret); return ret);

  // 5. (Boilerplate) Wait for task execution to complete synchronously.
  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  // 6. Obtain the output value and copy the result from device memory to host memory. Modify based on the specific API definition.
  auto size = GetShapeSize(outShape);
  std::vector<uint16_t> resultData(size, 0);
  ret = aclrtMemcpy(
      resultData.data(), resultData.size() * sizeof(resultData[0]), outDeviceAddr, size * sizeof(resultData[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);
  // fp16 data cannot be printed directly in C. Read it as uint16 and manually convert the binary representation to float-based fp16.
  for (int64_t i = 0; i < size; i++) {
      float fp16Float = Fp16ToFloat(resultData[i]);
      LOG_PRINT("result[%ld] is: %f\n", i, fp16Float);
  }

  // 7. Release aclTensor and aclScalar. Modify based on the specific API definition.
  aclDestroyTensor(self);
  aclDestroyTensor(mat2);
  aclDestroyTensor(out);

  // 8. Release device resources. Modify based on the specific API definition.
  aclrtFree(selfDeviceAddr);
  aclrtFree(mat2DeviceAddr);
  aclrtFree(outDeviceAddr);

  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  if (workspaceSizeMm > 0) {
    aclrtFree(workspaceAddrMm);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();
  return 0;
}