cuTensorNet Functions¶
Handle Management API¶
cutensornetCreate
¶
-
cutensornetStatus_t cutensornetCreate(cutensornetHandle_t *handle)¶
Initializes the cuTensorNet library.
The device associated with a particular cuTensorNet handle is assumed to remain unchanged after the cutensornetCreate() call. In order for the cuTensorNet library to use a different device, the application must set the new device to be used by calling cudaSetDevice() and then create another cuTensorNet handle, which will be associated with the new device, by calling cutensornetCreate().
Remark
blocking, no reentrant, and thread-safe
- Parameters
handle – [out] Pointer to cutensornetHandle_t
- Returns
CUTENSORNET_STATUS_SUCCESS on success and an error code otherwise
cutensornetDestroy
¶
-
cutensornetStatus_t cutensornetDestroy(cutensornetHandle_t handle)¶
Destroys the cuTensorNet library handle.
This function releases resources used by the cuTensorNet library handle. This function is the last call with a particular handle to the cuTensorNet library. Calling any cuTensorNet function which uses cutensornetHandle_t after cutensornetDestroy() will return an error.
- Parameters
handle – [inout] Opaque handle holding cuTensorNet’s library context.
Network Descriptor API¶
cutensornetCreateNetworkDescriptor
¶
-
cutensornetStatus_t cutensornetCreateNetworkDescriptor(const cutensornetHandle_t handle, int32_t numInputs, const int32_t numModesIn[], const int64_t *const extentsIn[], const int64_t *const stridesIn[], const int32_t *const modesIn[], const cutensornetTensorQualifiers_t qualifiersIn[], int32_t numModesOut, const int64_t extentsOut[], const int64_t stridesOut[], const int32_t modesOut[], cudaDataType_t dataType, cutensornetComputeType_t computeType, cutensornetNetworkDescriptor_t *descNet)¶
Initializes a cutensornetNetworkDescriptor_t, describing the connectivity (i.e., network topology) between the tensors.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyNetworkDescriptor() is called once
descNet
is no longer required.Supported data-type combinations are:
Data type
Compute type
Tensor Core
CUDA_R_16F
CUTENSORNET_COMPUTE_32F
Volta+
CUDA_R_16BF
CUTENSORNET_COMPUTE_32F
Ampere+
CUDA_R_32F
CUTENSORNET_COMPUTE_32F
No
CUDA_R_32F
CUTENSORNET_COMPUTE_TF32
Ampere+
CUDA_R_32F
CUTENSORNET_COMPUTE_16BF
Ampere+
CUDA_R_32F
CUTENSORNET_COMPUTE_16F
Volta+
CUDA_R_64F
CUTENSORNET_COMPUTE_64F
Ampere+
CUDA_R_64F
CUTENSORNET_COMPUTE_32F
No
CUDA_C_32F
CUTENSORNET_COMPUTE_32F
No
CUDA_C_32F
CUTENSORNET_COMPUTE_TF32
Ampere+
CUDA_C_64F
CUTENSORNET_COMPUTE_64F
Ampere+
CUDA_C_64F
CUTENSORNET_COMPUTE_32F
No
Note
If
stridesIn
(stridesOut
) is set to 0 (NULL
), it means the input tensors (output tensor) are in the Fortran (column-major) layout.numModesOut
can be set to-1
for cuTensorNet to infer the output modes based on the input modes, or to0
to perform a full reduction.If
qualifiersIn
is set to 0 (NULL
), cuTensorNet will use the defaults in cutensornetTensorQualifiers_t .- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
numInputs – [in] Number of input tensors.
numModesIn – [in] Array of size
numInputs
;numModesIn[i]
denotes the number of modes available in the i-th tensor.extentsIn – [in] Array of size
numInputs
;extentsIn[i]
hasnumModesIn[i]
many entries withextentsIn[i][j]
(j
<numModesIn[i]
) corresponding to the extent of the j-th mode of tensori
.stridesIn – [in] Array of size
numInputs
;stridesIn[i]
hasnumModesIn[i]
many entries withstridesIn[i][j]
(j
<numModesIn[i]
) corresponding to the linearized offset in physical memory between two logically-neighboring elements w.r.t the j-th mode of tensori
.modesIn – [in] Array of size
numInputs
;modesIn[i]
hasnumModesIn[i]
many entries each entry corresponds to a mode. Each mode that does not appear in the input tensor is implicitly contracted.qualifiersIn – [in] Array of size
numInputs
;qualifiersIn[i]
denotes the qualifiers of i-th input tensor. Refer to cutensornetTensorQualifiers_tnumModesOut – [in] number of modes of the output tensor. On entry, if this value is
-1
and the output modes are not provided, the network will infer the output modes. If this value is0
, the network is force reduced.extentsOut – [in] Array of size
numModesOut
;extentsOut[j]
(j
<numModesOut
) corresponding to the extent of the j-th mode of the output tensor.stridesOut – [in] Array of size
numModesOut
;stridesOut[j]
(j
<numModesOut
) corresponding to the linearized offset in physical memory between two logically-neighboring elements w.r.t the j-th mode of the output tensor.modesOut – [in] Array of size
numModesOut
;modesOut[j]
denotes the j-th mode of the output tensor. output tensor.dataType – [in] Denotes the data type for all input an output tensors.
computeType – [in] Denotes the compute type used throughout the computation.
descNet – [out] Pointer to a cutensornetNetworkDescriptor_t.
cutensornetDestroyNetworkDescriptor
¶
-
cutensornetStatus_t cutensornetDestroyNetworkDescriptor(cutensornetNetworkDescriptor_t desc)¶
Frees all the memory associated with the network descriptor.
- Parameters
desc – [inout] Opaque handle to a tensor network descriptor.
cutensornetNetworkGetAttribute
¶
-
cutensornetStatus_t cutensornetNetworkGetAttribute(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t networkDesc, cutensornetNetworkAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of networkDescriptor.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
networkDesc – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinnetworkDesc
.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetNetworkSetAttribute
¶
-
cutensornetStatus_t cutensornetNetworkSetAttribute(const cutensornetHandle_t handle, cutensornetNetworkDescriptor_t networkDesc, cutensornetNetworkAttributes_t attr, const void *buf, size_t sizeInBytes)¶
Sets attributes of networkDescriptor.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
networkDesc – [inout] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [in] This buffer (of size
sizeInBytes
) determines the value to whichattr
will be set.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetGetOutputTensorDetails
¶
-
cutensornetStatus_t cutensornetGetOutputTensorDetails(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, int32_t *numModes, size_t *dataSize, int32_t *modeLabels, int64_t *extents, int64_t *strides)¶
DEPRECATED: Gets the number of output modes, data size, modes, extents, and strides of the output tensor.
If all information regarding the output tensor is needed by the user, this function should be called twice (the first time to retrieve
numModesOut
for allocating memory, and the second to retrievemodesOut
,extentsOut
, andstridesOut
).- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Pointer to a cutensornetNetworkDescriptor_t.
numModes – [out] on return, holds the number of modes of the output tensor. Cannot be null.
dataSize – [out] if not null on return, holds the size (in bytes) of the memory needed for the output tensor. Optionally, can be null.
modeLabels – [out] if not null on return, holds the modes of the output tensor. Optionally, can be null.
extents – [out] if not null on return, holds the extents of the output tensor. Optionally, can be null.
strides – [out] if not null on return, holds the strides of the output tensor. Optionally, can be null.
cutensornetGetOutputTensorDescriptor
¶
-
cutensornetStatus_t cutensornetGetOutputTensorDescriptor(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, cutensornetTensorDescriptor_t *outputTensorDesc)¶
Creates a cutensornetTensorDescriptor_t representing the output tensor of the network.
This function will create a descriptor pointed to by
outputTensorDesc
. The user is responsible for calling cutensornetDestroyTensorDescriptor to destroy the descriptor.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Pointer to a cutensornetNetworkDescriptor_t.
outputTensorDesc – [out] an opaque cutensornetTensorDescriptor_t struct. Cannot be null. On return, a new cutensornetTensorDescriptor_t holds the meta-data of the
descNet
output tensor.
Tensor Descriptor API¶
cutensornetCreateTensorDescriptor
¶
-
cutensornetStatus_t cutensornetCreateTensorDescriptor(const cutensornetHandle_t handle, int32_t numModes, const int64_t extents[], const int64_t strides[], const int32_t modes[], cudaDataType_t dataType, cutensornetTensorDescriptor_t *descTensor)¶
Initializes a cutensornetTensorDescriptor_t, describing the information of a tensor.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyTensorDescriptor() is called once
descTensor
is no longer required.Note
If
strides
is set toNULL
, it means the tensor is in the Fortran (column-major) layout.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
numModes – [in] The number of modes of the tensor.
extents – [in] Array of size
numModes
;extents[j]
corresponding to the extent of the j-th mode of the tensor.strides – [in] Array of size
numModes
;strides[j]
corresponding to the linearized offset in physical memory between two logically-neighboring elements w.r.t the j-th mode of the tensor.modes – [in] Array of size
numModes
;modes[j]
denotes the j-th mode of the tensor.dataType – [in] Denotes the data type for the tensor.
descTensor – [out] Pointer to a cutensornetTensorDescriptor_t.
cutensornetGetTensorDetails
¶
-
cutensornetStatus_t cutensornetGetTensorDetails(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t tensorDesc, int32_t *numModes, size_t *dataSize, int32_t *modeLabels, int64_t *extents, int64_t *strides)¶
Gets the number of output modes, data size, mode labels, extents, and strides of a tensor.
If all information regarding the tensor is needed by the user, this function should be called twice (the first time to retrieve
numModes
for allocating memory, and the second to retrievemodeLabels
,extents
, andstrides
).- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
tensorDesc – [in] Opaque handle to a tensor descriptor.
numModes – [out] On return, holds the number of modes of the tensor. Cannot be null.
dataSize – [out] If not null on return, holds the size (in bytes) of the memory needed for the tensor. Optionally, can be null.
modeLabels – [out] If not null on return, holds the modes of the tensor. Optionally, can be null.
extents – [out] If not null on return, holds the extents of the tensor. Optionally, can be null.
strides – [out] If not null on return, holds the strides of the tensor. Optionally, can be null.
cutensornetDestroyTensorDescriptor
¶
-
cutensornetStatus_t cutensornetDestroyTensorDescriptor(cutensornetTensorDescriptor_t descTensor)¶
Frees all the memory associated with the tensor descriptor.
- Parameters
descTensor – [inout] Opaque handle to a tensor descriptor.
Contraction Optimizer API¶
cutensornetCreateContractionOptimizerConfig
¶
-
cutensornetStatus_t cutensornetCreateContractionOptimizerConfig(const cutensornetHandle_t handle, cutensornetContractionOptimizerConfig_t *optimizerConfig)¶
Sets up the required hyper-optimization parameters for the contraction order solver (see cutensornetContractionOptimize())
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyContractionOptimizerConfig() is called once
optimizerConfig
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
optimizerConfig – [out] This data structure holds all information about the user-requested hyper-optimization parameters.
cutensornetDestroyContractionOptimizerConfig
¶
-
cutensornetStatus_t cutensornetDestroyContractionOptimizerConfig(cutensornetContractionOptimizerConfig_t optimizerConfig)¶
Frees all the memory associated with
optimizerConfig
.- Parameters
optimizerConfig – [inout] Opaque structure.
cutensornetContractionOptimizerConfigGetAttribute
¶
-
cutensornetStatus_t cutensornetContractionOptimizerConfigGetAttribute(const cutensornetHandle_t handle, const cutensornetContractionOptimizerConfig_t optimizerConfig, cutensornetContractionOptimizerConfigAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of
optimizerConfig
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
optimizerConfig – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinoptimizerConfig
.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetContractionOptimizerConfigSetAttribute
¶
-
cutensornetStatus_t cutensornetContractionOptimizerConfigSetAttribute(const cutensornetHandle_t handle, cutensornetContractionOptimizerConfig_t optimizerConfig, cutensornetContractionOptimizerConfigAttributes_t attr, const void *buf, size_t sizeInBytes)¶
Sets attributes of
optimizerConfig
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
optimizerConfig – [inout] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [in] This buffer (of size
sizeInBytes
) determines the value to whichattr
will be set.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetCreateContractionOptimizerInfo
¶
-
cutensornetStatus_t cutensornetCreateContractionOptimizerInfo(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, cutensornetContractionOptimizerInfo_t *optimizerInfo)¶
Allocates resources for
optimizerInfo
.Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyContractionOptimizerInfo() is called once
optimizerInfo
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the tensor network (i.e., its tensors and their connectivity) for which
optimizerInfo
is created.optimizerInfo – [out] Pointer to cutensornetContractionOptimizerInfo_t.
cutensornetDestroyContractionOptimizerInfo
¶
-
cutensornetStatus_t cutensornetDestroyContractionOptimizerInfo(cutensornetContractionOptimizerInfo_t optimizerInfo)¶
Frees all the memory associated with
optimizerInfo
.- Parameters
optimizerInfo – [inout] Opaque structure.
cutensornetContractionOptimize
¶
-
cutensornetStatus_t cutensornetContractionOptimize(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, const cutensornetContractionOptimizerConfig_t optimizerConfig, uint64_t workspaceSizeConstraint, cutensornetContractionOptimizerInfo_t optimizerInfo)¶
Computes an “optimized” contraction order as well as slicing info (for more information see Overview section) for a given tensor network such that the total time to solution is minimized while adhering to the user-provided memory constraint.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the topology of the tensor network (i.e., all tensors, their connectivity and modes).
optimizerConfig – [in] Holds all hyper-optimization parameters that govern the search for an “optimal” contraction order.
workspaceSizeConstraint – [in] Maximal device memory that will be provided by the user (i.e., cuTensorNet has to find a viable path/slicing solution within this user-defined constraint).
optimizerInfo – [inout] On return, this object will hold all necessary information about the optimized path and the related slicing information.
optimizerInfo
will hold information including (see cutensornetContractionOptimizerInfoAttributes_t):Total number of slices.
Total number of sliced modes.
Information about the sliced modes (i.e., the IDs of the sliced modes (see
modesIn
w.r.t. cutensornetCreateNetworkDescriptor()) as well as their extents (see Overview section for additional documentation).Optimized path.
FLOP count.
Total number of elements in the largest intermediate tensor.
The mode labels for all intermediate tensors.
The estimated runtime and “effective” flops.
cutensornetContractionOptimizerInfoGetAttribute
¶
-
cutensornetStatus_t cutensornetContractionOptimizerInfoGetAttribute(const cutensornetHandle_t handle, const cutensornetContractionOptimizerInfo_t optimizerInfo, cutensornetContractionOptimizerInfoAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of
optimizerInfo
.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
optimizerInfo – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinoptimizeInfo
.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetContractionOptimizerInfoSetAttribute
¶
-
cutensornetStatus_t cutensornetContractionOptimizerInfoSetAttribute(const cutensornetHandle_t handle, cutensornetContractionOptimizerInfo_t optimizerInfo, cutensornetContractionOptimizerInfoAttributes_t attr, const void *buf, size_t sizeInBytes)¶
Sets attributes of optimizerInfo.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
optimizerInfo – [inout] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [in] This buffer (of size
sizeInBytes
) determines the value to whichattr
will be set.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetContractionOptimizerInfoGetPackedSize
¶
-
cutensornetStatus_t cutensornetContractionOptimizerInfoGetPackedSize(const cutensornetHandle_t handle, const cutensornetContractionOptimizerInfo_t optimizerInfo, size_t *sizeInBytes)¶
Gets the packed size of the
optimizerInfo
object.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
optimizerInfo – [in] Opaque structure of type cutensornetContractionOptimizerInfo_t.
sizeInBytes – [out] The packed size (in bytes).
cutensornetContractionOptimizerInfoPackData
¶
-
cutensornetStatus_t cutensornetContractionOptimizerInfoPackData(const cutensornetHandle_t handle, const cutensornetContractionOptimizerInfo_t optimizerInfo, void *buffer, size_t sizeInBytes)¶
Packs the
optimizerInfo
object into the provided buffer.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
optimizerInfo – [in] Opaque structure of type cutensornetContractionOptimizerInfo_t.
buffer – [out] On return, this buffer holds the contents of optimizerInfo in packed form.
sizeInBytes – [in] The size of the buffer (in bytes).
cutensornetCreateContractionOptimizerInfoFromPackedData
¶
-
cutensornetStatus_t cutensornetCreateContractionOptimizerInfoFromPackedData(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, const void *buffer, size_t sizeInBytes, cutensornetContractionOptimizerInfo_t *optimizerInfo)¶
Create an optimizerInfo object from the provided buffer.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the tensor network (i.e., its tensors and their connectivity) for which
optimizerInfo
is created.buffer – [in] A buffer with the contents of optimizerInfo in packed form.
sizeInBytes – [in] The size of the buffer (in bytes).
optimizerInfo – [out] Pointer to cutensornetContractionOptimizerInfo_t.
cutensornetUpdateContractionOptimizerInfoFromPackedData
¶
-
cutensornetStatus_t cutensornetUpdateContractionOptimizerInfoFromPackedData(const cutensornetHandle_t handle, const void *buffer, size_t sizeInBytes, cutensornetContractionOptimizerInfo_t optimizerInfo)¶
Update the provided
optimizerInfo
object from the provided buffer.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
buffer – [in] A buffer with the contents of optimizerInfo in packed form.
sizeInBytes – [in] The size of the buffer (in bytes).
optimizerInfo – [inout] Opaque object of type cutensornetContractionOptimizerInfo_t that will be updated.
Contraction Plan API¶
cutensornetCreateContractionPlan
¶
-
cutensornetStatus_t cutensornetCreateContractionPlan(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, const cutensornetContractionOptimizerInfo_t optimizerInfo, const cutensornetWorkspaceDescriptor_t workDesc, cutensornetContractionPlan_t *plan)¶
Initializes a cutensornetContractionPlan_t.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyContractionPlan() is called once
plan
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the tensor network (i.e., its tensors and their connectivity).
optimizerInfo – [in] Opaque structure.
workDesc – [in] Opaque structure describing the workspace. At the creation of the contraction plan, only the workspace size is needed; the pointer to the workspace memory may be left null. If a device memory handler is set,
workDesc
can be set either to null (in which case the “recommended” workspace size is inferred, see CUTENSORNET_WORKSIZE_PREF_RECOMMENDED) or to a valid cutensornetWorkspaceDescriptor_t with the desired workspace size set and a null workspace pointer, see Memory Management API section.plan – [out] cuTensorNet’s contraction plan holds all the information required to perform the tensor contractions; to be precise, it initializes a
cutensorContractionPlan_t
for each tensor contraction that is required to contract the entire tensor network.
cutensornetDestroyContractionPlan
¶
-
cutensornetStatus_t cutensornetDestroyContractionPlan(cutensornetContractionPlan_t plan)¶
Frees all resources owned by
plan
.- Parameters
plan – [inout] Opaque structure.
cutensornetContractionAutotune
¶
-
cutensornetStatus_t cutensornetContractionAutotune(const cutensornetHandle_t handle, cutensornetContractionPlan_t plan, const void *const rawDataIn[], void *rawDataOut, cutensornetWorkspaceDescriptor_t workDesc, const cutensornetContractionAutotunePreference_t pref, cudaStream_t stream)¶
Auto-tunes the contraction plan to find the best
cutensorContractionPlan_t
for each pair-wise contraction.Note
This function is blocking due to the nature of the auto-tuning process.
Input and output data pointers are recommended to be 256-byte aligned for best performance.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
plan – [inout] The plan must already be created (see cutensornetCreateContractionPlan()); the individual contraction plans will be fine-tuned.
rawDataIn – [in] Array of N pointers (N being the number of input tensors specified cutensornetCreateNetworkDescriptor());
rawDataIn[i]
points to the data associated with the i-th input tensor (in device memory).rawDataOut – [out] Points to the raw data of the output tensor (in device memory).
workDesc – [in] Opaque structure describing the workspace. The provided workspace must be valid (the workspace size must be the same as or larger than both the minimum needed and the value provided at plan creation). See cutensornetCreateContractionPlan(), cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory(). If a device memory handler is set, the
workDesc
can be set to null, or the workspace pointer inworkDesc
can be set to null, and the workspace size can be set either to 0 (in which case the “recommended” size is used, see CUTENSORNET_WORKSIZE_PREF_RECOMMENDED) or to a valid size. A workspace of the specified size will be drawn from the user’s mempool and released back once done.pref – [in] Controls the auto-tuning process and gives the user control over how much time is spent in this routine.
stream – [in] The CUDA stream on which the computation is performed.
cutensornetCreateContractionAutotunePreference
¶
-
cutensornetStatus_t cutensornetCreateContractionAutotunePreference(const cutensornetHandle_t handle, cutensornetContractionAutotunePreference_t *autotunePreference)¶
Sets up the required auto-tune parameters for the contraction plan.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyContractionAutotunePreference() is called once
autotunePreference
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
autotunePreference – [out] This data structure holds all information about the user-requested auto-tune parameters.
cutensornetContractionAutotunePreferenceGetAttribute
¶
-
cutensornetStatus_t cutensornetContractionAutotunePreferenceGetAttribute(const cutensornetHandle_t handle, const cutensornetContractionAutotunePreference_t autotunePreference, cutensornetContractionAutotunePreferenceAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of
autotunePreference
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
autotunePreference – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinautotunePreference
.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetContractionAutotunePreferenceSetAttribute
¶
-
cutensornetStatus_t cutensornetContractionAutotunePreferenceSetAttribute(const cutensornetHandle_t handle, cutensornetContractionAutotunePreference_t autotunePreference, cutensornetContractionAutotunePreferenceAttributes_t attr, const void *buf, size_t sizeInBytes)¶
Sets attributes of
autotunePreference
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
autotunePreference – [inout] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [in] This buffer (of size
sizeInBytes
) determines the value to whichattr
will be set.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetDestroyContractionAutotunePreference
¶
-
cutensornetStatus_t cutensornetDestroyContractionAutotunePreference(cutensornetContractionAutotunePreference_t autotunePreference)¶
Frees all the memory associated with
autotunePreference
.- Parameters
autotunePreference – [inout] Opaque structure.
Workspace Management API¶
cutensornetCreateWorkspaceDescriptor
¶
-
cutensornetStatus_t cutensornetCreateWorkspaceDescriptor(const cutensornetHandle_t handle, cutensornetWorkspaceDescriptor_t *workDesc)¶
Creates a workspace descriptor that holds information about the user provided memory buffer.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [out] Pointer to the opaque workspace descriptor.
cutensornetWorkspaceComputeSizes
¶
-
cutensornetStatus_t cutensornetWorkspaceComputeSizes(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, const cutensornetContractionOptimizerInfo_t optimizerInfo, cutensornetWorkspaceDescriptor_t workDesc)¶
DEPRECATED: Computes the workspace size needed to contract the input tensor network using the provided contraction path.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the tensor network (i.e., its tensors and their connectivity).
optimizerInfo – [in] Opaque structure.
workDesc – [out] The workspace descriptor in which the information is collected.
cutensornetWorkspaceComputeContractionSizes
¶
-
cutensornetStatus_t cutensornetWorkspaceComputeContractionSizes(const cutensornetHandle_t handle, const cutensornetNetworkDescriptor_t descNet, const cutensornetContractionOptimizerInfo_t optimizerInfo, cutensornetWorkspaceDescriptor_t workDesc)¶
Computes the workspace size needed to contract the input tensor network using the provided contraction path.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descNet – [in] Describes the tensor network (i.e., its tensors and their connectivity).
optimizerInfo – [in] Opaque structure.
workDesc – [out] The workspace descriptor in which the information is collected.
cutensornetWorkspaceComputeQRSizes
¶
-
cutensornetStatus_t cutensornetWorkspaceComputeQRSizes(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorIn, const cutensornetTensorDescriptor_t descTensorQ, const cutensornetTensorDescriptor_t descTensorR, cutensornetWorkspaceDescriptor_t workDesc)¶
Computes the workspace size needed to perform the tensor QR operation.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorIn – [in] Describes the modes, extents and other metadata information for a tensor.
descTensorQ – [in] Describes the modes, extents and other metadata information for the output tensor Q.
descTensorR – [in] Describes the modes, extents and other metadata information for the output tensor R.
workDesc – [out] The workspace descriptor in which the information is collected.
cutensornetWorkspaceComputeSVDSizes
¶
-
cutensornetStatus_t cutensornetWorkspaceComputeSVDSizes(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorIn, const cutensornetTensorDescriptor_t descTensorU, const cutensornetTensorDescriptor_t descTensorV, const cutensornetTensorSVDConfig_t svdConfig, cutensornetWorkspaceDescriptor_t workDesc)¶
Computes the workspace size needed to perform the tensor SVD operation.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorIn – [in] Describes the modes, extents and other metadata information for a tensor.
descTensorU – [in] Describes the modes, extents and other metadata information for the output tensor U.
descTensorV – [in] Describes the modes, extents and other metadata information for the output tensor V.
svdConfig – [in] This data structure holds the user-requested svd parameters.
workDesc – [out] The workspace descriptor in which the information is collected.
cutensornetWorkspaceComputeGateSplitSizes
¶
-
cutensornetStatus_t cutensornetWorkspaceComputeGateSplitSizes(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorInA, const cutensornetTensorDescriptor_t descTensorInB, const cutensornetTensorDescriptor_t descTensorInG, const cutensornetTensorDescriptor_t descTensorU, const cutensornetTensorDescriptor_t descTensorV, const cutensornetGateSplitAlgo_t gateAlgo, const cutensornetTensorSVDConfig_t svdConfig, cutensornetComputeType_t computeType, cutensornetWorkspaceDescriptor_t workDesc)¶
Computes the workspace size needed to perform the gating operation.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorInA – [in] Describes the modes, extents, and other metadata information of the input tensor A.
descTensorInB – [in] Describes the modes, extents, and other metadata information of the input tensor B.
descTensorInG – [in] Describes the modes, extents, and other metadata information of the input gate tensor.
descTensorU – [in] Describes the modes, extents, and other metadata information of the output U tensor. The extents of uncontracted modes are expected to be consistent with
descTensorInA
anddescTensorInG
.descTensorV – [in] Describes the modes, extents, and other metadata information of the output V tensor. The extents of uncontracted modes are expected to be consistent with
descTensorInB
anddescTensorInG
.gateAlgo – [in] The algorithm to use for splitting the gate tensor onto tensor A and B.
svdConfig – [in] Opaque structure holding the user-requested SVD parameters.
computeType – [in] Denotes the compute type used throughout the computation.
workDesc – [out] Opaque structure describing the workspace.
cutensornetWorkspaceGetSize
¶
-
cutensornetStatus_t cutensornetWorkspaceGetSize(const cutensornetHandle_t handle, const cutensornetWorkspaceDescriptor_t workDesc, cutensornetWorksizePref_t workPref, cutensornetMemspace_t memSpace, uint64_t *workspaceSize)¶
DEPRECATED: Retrieves the needed workspace size for the given workspace preference and memory space.
The needed sizes for different tasks must be pre-calculated by calling the corresponding API, e.g, cutensornetWorkspaceComputeContractionSizes(), cutensornetWorkspaceComputeQRSizes(), cutensornetWorkspaceComputeSVDSizes() and cutensornetWorkspaceComputeGateSplitSizes().
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [in] Opaque structure describing the workspace.
workPref – [in] Preference of workspace for planning.
memSpace – [in] The memory space where the workspace is allocated.
workspaceSize – [out] Needed workspace size.
cutensornetWorkspaceGetMemorySize
¶
-
cutensornetStatus_t cutensornetWorkspaceGetMemorySize(const cutensornetHandle_t handle, const cutensornetWorkspaceDescriptor_t workDesc, cutensornetWorksizePref_t workPref, cutensornetMemspace_t memSpace, cutensornetWorkspaceKind_t workKind, int64_t *memorySize)¶
Retrieves the needed workspace size for the given workspace preference, memory space, workspace kind.
The needed sizes for different tasks must be pre-calculated by calling the corresponding API, e.g, cutensornetWorkspaceComputeContractionSizes(), cutensornetWorkspaceComputeQRSizes(), cutensornetWorkspaceComputeSVDSizes() and cutensornetWorkspaceComputeGateSplitSizes().
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [in] Opaque structure describing the workspace.
workPref – [in] Preference of workspace for planning.
memSpace – [in] The memory space where the workspace is allocated.
workKind – [in] The kind of workspace.
memorySize – [out] Needed workspace size.
cutensornetWorkspaceSet
¶
-
cutensornetStatus_t cutensornetWorkspaceSet(const cutensornetHandle_t handle, cutensornetWorkspaceDescriptor_t workDesc, cutensornetMemspace_t memSpace, void *const workspacePtr, uint64_t workspaceSize)¶
DEPRECATED: Sets the memory address and workspace size of workspace provided by user.
A workspace is valid in the following cases:
workspacePtr
is valid andworkspaceSize
> 0workspacePtr
is null andworkspaceSize
> 0 (used during cutensornetCreateContractionPlan() to provide the available workspace).workspacePtr
is null andworkspaceSize
= 0 (workspace memory will be drawn from the user’s mempool)
A workspace will be validated against the minimal required at usage (cutensornetCreateContractionPlan(), cutensornetContractionAutotune(), cutensornetContraction())
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [inout] Opaque structure describing the workspace.
memSpace – [in] The memory space where the workspace is allocated.
workspacePtr – [in] Workspace memory pointer, may be null.
workspaceSize – [in] Workspace size, must be >= 0.
cutensornetWorkspaceSetMemory
¶
-
cutensornetStatus_t cutensornetWorkspaceSetMemory(const cutensornetHandle_t handle, cutensornetWorkspaceDescriptor_t workDesc, cutensornetMemspace_t memSpace, cutensornetWorkspaceKind_t workKind, void *const memoryPtr, int64_t memorySize)¶
Sets the memory address and workspace size of the workspace provided by user.
A workspace is valid in the following cases:
memoryPtr
is valid andmemorySize
> 0memoryPtr
is null andmemorySize
> 0: used to indicate memory with the indicatedmemorySize
should be drawn from the mempool, or for cutensornetCreateContractionPlan() to indicate the available workspace size.memoryPtr
is null andmemorySize
= 0: indicates the workspace of the specified kind is disabled (currently applies to CACHE kind only).memoryPtr
is null andmemorySize
< 0: indicates workspace memory should be drawn from the user’s mempool with the CUTENSORNET_WORKSIZE_PREF_RECOMMENDED size (see cutensornetWorksizePref_t).
The
memorySize
of the SCRATCH kind will be validated against the minimal required at usage (cutensornetCreateContractionPlan(), cutensornetContractionAutotune(), cutensornetContraction(), cutensornetContractSlices()) The CACHE memory size can be any, the larger the better.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [inout] Opaque structure describing the workspace.
memSpace – [in] The memory space where the workspace is allocated.
workKind – [in] The kind of workspace.
memoryPtr – [in] Workspace memory pointer, may be null.
memorySize – [in] Workspace size.
cutensornetWorkspaceGet
¶
-
cutensornetStatus_t cutensornetWorkspaceGet(const cutensornetHandle_t handle, const cutensornetWorkspaceDescriptor_t workDesc, cutensornetMemspace_t memSpace, void **workspacePtr, uint64_t *workspaceSize)¶
DEPRECATED: Retrieves the memory address and workspace size of workspace hosted in the workspace descriptor.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [in] Opaque structure describing the workspace.
memSpace – [in] The memory space where the workspace is allocated.
workspacePtr – [out] Workspace memory pointer.
workspaceSize – [out] Workspace size.
cutensornetWorkspaceGetMemory
¶
-
cutensornetStatus_t cutensornetWorkspaceGetMemory(const cutensornetHandle_t handle, const cutensornetWorkspaceDescriptor_t workDesc, cutensornetMemspace_t memSpace, cutensornetWorkspaceKind_t workKind, void **memoryPtr, int64_t *memorySize)¶
Retrieves the memory address and workspace size of workspace hosted in the workspace descriptor.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [in] Opaque structure describing the workspace.
memSpace – [in] The memory space where the workspace is allocated.
workKind – [in] The kind of workspace.
memoryPtr – [out] Workspace memory pointer.
memorySize – [out] Workspace size.
cutensornetWorkspacePurgeCache
¶
-
cutensornetStatus_t cutensornetWorkspacePurgeCache(const cutensornetHandle_t handle, cutensornetWorkspaceDescriptor_t workDesc, cutensornetMemspace_t memSpace)¶
Purges the cached data in the specified memory space.
Purges/invalidates the cached data in the CUTENSORNET_WORKSPACE_CACHE workspace kind on the
memSpace
memory space, but does not free the memory nor return it to the memory pool.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
workDesc – [inout] Opaque structure describing the workspace.
memSpace – [in] The memory space where the workspace is allocated.
cutensornetDestroyWorkspaceDescriptor
¶
-
cutensornetStatus_t cutensornetDestroyWorkspaceDescriptor(cutensornetWorkspaceDescriptor_t desc)¶
Frees the workspace descriptor.
Note that this API does not free the memory provided by cutensornetWorkspaceSetMemory().
- Parameters
desc – [inout] Opaque structure.
Network Contraction API¶
cutensornetContraction
¶
-
cutensornetStatus_t cutensornetContraction(const cutensornetHandle_t handle, cutensornetContractionPlan_t plan, const void *const rawDataIn[], void *rawDataOut, cutensornetWorkspaceDescriptor_t workDesc, int64_t sliceId, cudaStream_t stream)¶
DEPRECATED: Performs the actual contraction of the tensor network.
Note
If multiple slices are created, the order of contracting over slices using cutensornetContraction() should be ascending starting from slice 0. If parallelizing over slices manually (in any fashion: streams, devices, processes, etc.), please make sure the output tensors (that are subject to a global reduction) are zero-initialized.
Input and output data pointers are recommended to be 256-byte aligned for best performance.
This function is asynchronous w.r.t. the calling CPU thread. The user should guarantee that the memory buffer provided in
workDesc
is valid until a synchronization with the stream or the device is executed.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
plan – [inout] Encodes the execution of a tensor network contraction (see cutensornetCreateContractionPlan() and cutensornetContractionAutotune()). Some internal meta-data may be updated upon contraction.
rawDataIn – [in] Array of N pointers (N being the number of input tensors specified cutensornetCreateNetworkDescriptor());
rawDataIn[i]
points to the data associated with the i-th input tensor (in device memory).rawDataOut – [out] Points to the raw data of the output tensor (in device memory).
workDesc – [in] Opaque structure describing the workspace. The provided workspace must be valid (the workspace size must be the same as or larger than both the minimum needed and the value provided at plan creation). See cutensornetCreateContractionPlan(), cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory()). If a device memory handler is set, then
workDesc
can be set to null, or the workspace pointer inworkDesc
can be set to null, and the workspace size can be set either to 0 (in which case the “recommended” size is used, see CUTENSORNET_WORKSIZE_PREF_RECOMMENDED) or to a valid size. A workspace of the specified size will be drawn from the user’s mempool and released back once done.sliceId – [in] The ID of the slice that is currently contracted (this value ranges between
0
andoptimizerInfo.numSlices
); use0
if no slices are used.stream – [in] The CUDA stream on which the computation is performed.
cutensornetContractSlices
¶
-
cutensornetStatus_t cutensornetContractSlices(const cutensornetHandle_t handle, cutensornetContractionPlan_t plan, const void *const rawDataIn[], void *rawDataOut, int32_t accumulateOutput, cutensornetWorkspaceDescriptor_t workDesc, const cutensornetSliceGroup_t sliceGroup, cudaStream_t stream)¶
Performs the actual contraction of the tensor network.
Note
Input and output data pointers are recommended to be at least 256-byte aligned for best performance.
Warning
In the current release, this function will synchronize the stream in case distributed execution is activated (via cutensornetDistributedResetConfiguration)
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
plan – [inout] Encodes the execution of a tensor network contraction (see cutensornetCreateContractionPlan() and cutensornetContractionAutotune()). Some internal meta-data may be updated upon contraction.
rawDataIn – [in] Array of N pointers (N being the number of input tensors specified in cutensornetCreateNetworkDescriptor()):
rawDataIn[i]
points to the data associated with the i-th input tensor (in device memory).rawDataOut – [out] Points to the raw data of the output tensor (in device memory).
accumulateOutput – [in] If 0, write the contraction result into rawDataOut; otherwise accumulate the result into rawDataOut.
workDesc – [in] Opaque structure describing the workspace. The provided CUTENSORNET_WORKSPACE_SCRATCH workspace must be valid (the workspace size must be the same as or larger than both the minimum needed and the value provided at plan creation). See cutensornetCreateContractionPlan(), cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory(). The provided CUTENSORNET_WORKSPACE_CACHE workspace can be of any size, the larger the better, up to the size that can be queried with cutensornetWorkspaceGetMemorySize(). If a device memory handler is set, then
workDesc
can be set to null, or the memory pointer inworkDesc
of either the workspace kinds can be set to null, and the workspace size can be set either to a negative value (in which case the “recommended” size is used, see CUTENSORNET_WORKSIZE_PREF_RECOMMENDED) or to a valid size. For a workspace of kind CUTENSORNET_WORKSPACE_SCRATCH, a memory buffer with the specified size will be drawn from the user’s mempool and released back once done. For a workspace of kind CUTENSORNET_WORKSPACE_CACHE, a memory buffer with the specified size will be drawn from the user’s mempool and released back once theworkDesc
is destroyed, ifworkDesc
!= NULL, otherwise, once theplan
is destroyed, or an alternativeworkDesc
with a different memory address/size is provided in a subsequent cutensornetContractSlices() call.sliceGroup – [in] Opaque object specifying the slices to be contracted (see cutensornetCreateSliceGroupFromIDRange() and cutensornetCreateSliceGroupFromIDs()). If set to null, all slices will be contracted.
stream – [in] The CUDA stream on which the computation is performed.
Gradient Computation API¶
cutensornetComputeGradientsBackward
¶
-
cutensornetStatus_t cutensornetComputeGradientsBackward(const cutensornetHandle_t handle, cutensornetContractionPlan_t plan, const void *const rawDataIn[], const void *outputGradient, void *const gradients[], int32_t accumulateOutput, cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t stream)¶
Computes the gradients of the network w.r.t. the input tensors whose gradients are required. The network must have been contracted and loaded in the
workDesc
CACHE. Operates only on networks with single slice and no singleton modes.Note
This function is experimental and is subject to change in future releases.
Note
This function should be preceded with a call to cutensornetContractSlices(); Both calls to cutensornetContractSlices() and cutensornetComputeGradientsBackward() should use either the same
workDesc
instance (in order to share the CACHE memory), or both pass null toworkDesc
to use same mempool allocation for CACHE.workDesc
andplan
should not be altered in between these calls.Calling cutensornetWorkspacePurgeCache() is necessary for computing gradients of different data sets (the combo cutensornetContractSlices() and cutensornetComputeGradientsBackward() calls generate cached data that is only valid for the corresponding dataset and should be purged when the input tensors’ data change)
Input data’s, output data’s, and workspace buffers’ pointers are recommended to be at least 256-byte aligned for best performance.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
plan – [inout] Encodes the execution of a tensor network contraction (see cutensornetCreateContractionPlan() and cutensornetContractionAutotune()). Some internal meta-data may be updated upon contraction.
rawDataIn – [in] Array of N pointers (N being the number of input tensors specified in cutensornetCreateNetworkDescriptor()):
rawDataIn[i]
points to the data associated with the i-th input tensor (in device memory).outputGradient – [in] Gradient of the output tensor (in device memory). Must have the same memory layout (strides) as the output tensor of the tensor network.
gradients – [inout] Array of N pointers:
gradients[i]
points to the gradient data associated with the i-th input tensor in device memory. Settinggradients[i]
to null would skip computing the gradient of the i-th input tensor. Generated gradient data has the same memory layout (strides) as their corresponding input tensors.accumulateOutput – [in] If 0, write the gradient results into
gradients
; otherwise accumulates the results intogradients
.workDesc – [in] Opaque structure describing the workspace. The provided CUTENSORNET_WORKSPACE_SCRATCH workspace must be valid (the workspace size must be the same as or larger than the minimum needed). See cutensornetWorkspaceComputeContractionSizes(), cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory(). The provided CUTENSORNET_WORKSPACE_CACHE workspace must be valid, and contains the cached intermediate tensors from the corresponding cutensornetContractSlices() call. If a device memory handler is set, and
workDesc
is set to null, or the memory pointer inworkDesc
of either the workspace kinds is set to null, for both calls to cutensornetContractSlices() and cutensornetComputeGradientsBackward(), memory will be drawn from the memory pool. See cutensornetContractSlices() for details.stream – [in] The CUDA stream on which the computation is performed.
Slice Group API¶
cutensornetCreateSliceGroupFromIDRange
¶
-
cutensornetStatus_t cutensornetCreateSliceGroupFromIDRange(const cutensornetHandle_t handle, int64_t sliceIdStart, int64_t sliceIdStop, int64_t sliceIdStep, cutensornetSliceGroup_t *sliceGroup)¶
Create a
cutensornetSliceGroup_t
object from a range, which produces a sequence of slice IDs from the specified start (inclusive) to the specified stop (exclusive) values with the specified step. The sequence can be increasing or decreasing depending on the start and stop values.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
sliceIdStart – [in] The start slice ID.
sliceIdStop – [in] The final slice ID is the largest (smallest) integer that excludes this value and all those above (below) for an increasing (decreasing) sequence.
sliceIdStep – [in] The step size between two successive slice IDs. A negative step size should be specified for a decreasing sequence.
sliceGroup – [out] Opaque object specifying the slice IDs.
cutensornetCreateSliceGroupFromIDs
¶
-
cutensornetStatus_t cutensornetCreateSliceGroupFromIDs(const cutensornetHandle_t handle, const int64_t *beginIDSequence, const int64_t *endIDSequence, cutensornetSliceGroup_t *sliceGroup)¶
Create a
cutensornetSliceGroup_t
object from a sequence of slice IDs. Duplicates in the input slice ID sequence will be removed.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
beginIDSequence – [in] A pointer to the beginning of the slice ID sequence.
endIDSequence – [in] A pointer to the end of the slice ID sequence.
sliceGroup – [out] Opaque object specifying the slice IDs.
cutensornetDestroySliceGroup
¶
-
cutensornetStatus_t cutensornetDestroySliceGroup(cutensornetSliceGroup_t sliceGroup)¶
Releases the resources associated with a
cutensornetSliceGroup_t
object and sets its value to null.- Parameters
sliceGroup – [inout] Opaque object specifying the slices to be contracted (see cutensornetCreateSliceGroupFromIDRange() and cutensornetCreateSliceGroupFromIDs()).
Approximate Tensor Network Execution API¶
cutensornetTensorQR
¶
-
cutensornetStatus_t cutensornetTensorQR(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorIn, const void *const rawDataIn, const cutensornetTensorDescriptor_t descTensorQ, void *q, const cutensornetTensorDescriptor_t descTensorR, void *r, const cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t stream)¶
Performs QR decomposition of a tensor.
The partition of all input modes in
descTensorIn
is specified indescTensorQ
anddescTensorR
.descTensorQ
anddescTensorR
are expected to share exactly one mode and the extent of that mode shall not exceed the minimum of m (row dimension) and n (column dimension) of the equivalent combined matrix QR.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorIn – [in] Describes the modes, extents, and other metadata information of a tensor.
rawDataIn – [in] Pointer to the raw data of the input tensor (in device memory).
descTensorQ – [in] Describes the modes, extents, and other metadata information of the output tensor Q.
q – [out] Pointer to the output tensor data Q (in device memory).
descTensorR – [in] Describes the modes, extents, and other metadata information of the output tensor R.
r – [out] Pointer to the output tensor data R (in device memory).
workDesc – [in] Opaque structure describing the workspace. The provided workspace must be valid (the workspace size must be the same as or larger than the minimum needed). See cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory().
stream – [in] The CUDA stream on which the computation is performed.
cutensornetTensorSVD
¶
-
cutensornetStatus_t cutensornetTensorSVD(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorIn, const void *const rawDataIn, cutensornetTensorDescriptor_t descTensorU, void *u, void *s, cutensornetTensorDescriptor_t descTensorV, void *v, const cutensornetTensorSVDConfig_t svdConfig, cutensornetTensorSVDInfo_t svdInfo, const cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t stream)¶
Performs SVD decomposition of a tensor.
The partition of all input modes in
descTensorIn
is specified indescTensorU
anddescTensorV
.descTensorU
anddescTensorV
are expected to share exactly one mode. The extent of the shared mode shall not exceed the minimum of m (row dimension) and n (column dimension) for the equivalent combined matrix SVD. The following variants of tensor SVD are supported:1. Exact SVD: This can be specified by setting the extent of the shared mode in
descTensorU
anddescTensorV
to be the mininum of m and n, and settingsvdConfig
toNULL
.2. SVD with fixed extent truncation: This can be specified by setting the extent of the shared mode in
descTensorU
anddescTensorV
to be lower than the mininum of m and n.3. SVD with value-based truncation: This can be specified by setting CUTENSORNET_TENSOR_SVD_CONFIG_ABS_CUTOFF or CUTENSORNET_TENSOR_SVD_CONFIG_REL_CUTOFF attribute of
svdConfig
.4. SVD with a combination of fixed extent and value-based truncation as described above.
Note
In the case of exact SVD or SVD with fixed extent truncation,
descTensorU
anddescTensorV
will remain constant after the execution. The data inu
andv
will respect theextent
andstride
in these tensor descriptors.When value-based truncation is requested in
svdConfig
,cutensornetTensorSVD
searches for the minimal extent that satifies both the value-based truncation and fixed extent requirement. If the resulting extent is found to be the same as the one specified in U/V tensor descriptors, theextent
andstride
from the tensor descriptors will be respected. If the resulting extent is found to be lower than the one specified in U/V tensor descriptors, the data inu
andv
will adopt a new Fortran-layout matching the reduced extent found. Theextent
andstride
indescTensorU
anddescTensorV
will also be overwritten to reflect this change. The user can query the reduced extent with cutensornetTensorSVDInfoGetAttribute() or cutensornetGetTensorDetails() (which also returns the new strides).As the reduced size for value-based truncation is not known until runtime, the user should always allocate based on the full data size specified by the initial
descTensorU
anddescTensorV
foru
andv
.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorIn – [in] Describes the modes, extents, and other metadata information of a tensor.
rawDataIn – [in] Pointer to the raw data of the input tensor (in device memory).
descTensorU – [inout] Describes the modes, extents, and other metadata information of the output tensor U. The extents for uncontracted modes are expected to be consistent with
descTensorIn
.u – [out] Pointer to the output tensor data U (in device memory).
s – [out] Pointer to the output tensor data S (in device memory). Can be
NULL
when the CUTENSORNET_TENSOR_SVD_CONFIG_S_PARTITION attribute ofsvdConfig
is not set to default (CUTENSORNET_TENSOR_SVD_PARTITION_NONE).descTensorV – [inout] Describes the modes, extents, and other metadata information of the output tensor V.
v – [out] Pointer to the output tensor data V (in device memory).
svdConfig – [in] This data structure holds the user-requested SVD parameters. Can be
NULL
if users do not need to perform value-based truncation or singular value partitioning.svdInfo – [out] Opaque structure holding all information about the trucation at runtime. Can be
NULL
if runtime information on singular value truncation is not needed.workDesc – [in] Opaque structure describing the workspace. The provided workspace must be valid (the workspace size must be the same as or larger than the minimum needed). See cutensornetWorkspaceGetMemorySize() & cutensornetWorkspaceSetMemory().
stream – [in] The CUDA stream on which the computation is performed.
cutensornetGateSplit
¶
-
cutensornetStatus_t cutensornetGateSplit(const cutensornetHandle_t handle, const cutensornetTensorDescriptor_t descTensorInA, const void *rawDataInA, const cutensornetTensorDescriptor_t descTensorInB, const void *rawDataInB, const cutensornetTensorDescriptor_t descTensorInG, const void *rawDataInG, cutensornetTensorDescriptor_t descTensorU, void *u, void *s, cutensornetTensorDescriptor_t descTensorV, void *v, const cutensornetGateSplitAlgo_t gateAlgo, const cutensornetTensorSVDConfig_t svdConfig, cutensornetComputeType_t computeType, cutensornetTensorSVDInfo_t svdInfo, const cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t stream)¶
Performs gate split operation.
descTensorInA
,descTensorInB
, anddescTensorInG
are expected to form a fully connected graph where the uncontracted modes are partitioned ontodescTensorU
anddescTensorV
via tensor SVD.descTensorU
anddescTensorV
are expected to share exactly one mode. The extent of that mode shall not exceed the minimum of m (row dimension) and n (column dimension) of the smallest equivalent matrix SVD problem.Note
The options for truncation and the treatment of
extent
andstride
follows the same logic as tensor SVD, see cutensornetTensorSVD().- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
descTensorInA – [in] Describes the modes, extents, and other metadata information of the input tensor A.
rawDataInA – [in] Pointer to the raw data of the input tensor A (in device memory).
descTensorInB – [in] Describes the modes, extents, and other metadata information of the input tensor B.
rawDataInB – [in] Pointer to the raw data of the input tensor B (in device memory).
descTensorInG – [in] Describes the modes, extents, and other metadata information of the input gate tensor.
rawDataInG – [in] Pointer to the raw data of the input gate tensor G (in device memory).
descTensorU – [in] Describes the modes, extents, and other metadata information of the output U tensor. The extents of uncontracted modes are expected to be consistent with
descTensorInA
anddescTensorInG
.u – [out] Pointer to the output tensor data U (in device memory).
s – [out] Pointer to the output tensor data S (in device memory). Can be
NULL
when the CUTENSORNET_TENSOR_SVD_CONFIG_S_PARTITION attribute ofsvdConfig
is not set to default (CUTENSORNET_TENSOR_SVD_PARTITION_NONE).descTensorV – [in] Describes the modes, extents, and other metadata information of the output V tensor. The extents of uncontracted modes are expected to be consistent with
descTensorInB
anddescTensorInG
.v – [out] Pointer to the output tensor data V (in device memory).
gateAlgo – [in] The algorithm to use for splitting the gate tensor into tensor A and B.
svdConfig – [in] Opaque structure holding the user-requested SVD parameters.
computeType – [in] Denotes the compute type used throughout the computation.
svdInfo – [out] Opaque structure holding all information about the truncation at runtime.
workDesc – [in] Opaque structure describing the workspace.
stream – [in] The CUDA stream on which the computation is performed.
Tensor SVD Config API¶
cutensornetCreateTensorSVDConfig
¶
-
cutensornetStatus_t cutensornetCreateTensorSVDConfig(const cutensornetHandle_t handle, cutensornetTensorSVDConfig_t *svdConfig)¶
Sets up the options for singular value decomposition and truncation.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyTensorSVDConfig() is called once
svdConfig
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
svdConfig – [out] This data structure holds the user-requested svd parameters.
cutensornetDestroyTensorSVDConfig
¶
-
cutensornetStatus_t cutensornetDestroyTensorSVDConfig(cutensornetTensorSVDConfig_t svdConfig)¶
Frees all the memory associated with the tensor svd configuration.
- Parameters
svdConfig – [inout] Opaque handle to a tensor svd configuration.
cutensornetTensorSVDConfigGetAttribute
¶
-
cutensornetStatus_t cutensornetTensorSVDConfigGetAttribute(const cutensornetHandle_t handle, const cutensornetTensorSVDConfig_t svdConfig, cutensornetTensorSVDConfigAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of
svdConfig
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
svdConfig – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinsvdConfig
.sizeInBytes – [in] Size of
buf
(in bytes).
cutensornetTensorSVDConfigSetAttribute
¶
-
cutensornetStatus_t cutensornetTensorSVDConfigSetAttribute(const cutensornetHandle_t handle, cutensornetTensorSVDConfig_t svdConfig, cutensornetTensorSVDConfigAttributes_t attr, const void *buf, size_t sizeInBytes)¶
Sets attributes of
svdConfig
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
svdConfig – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [in] This buffer (of size
sizeInBytes
) determines the value to whichattr
will be set.sizeInBytes – [in] Size of
buf
(in bytes).
Tensor SVD Info API¶
cutensornetCreateTensorSVDInfo
¶
-
cutensornetStatus_t cutensornetCreateTensorSVDInfo(const cutensornetHandle_t handle, cutensornetTensorSVDInfo_t *svdInfo)¶
Sets up the information for singular value decomposition.
Note that this function allocates data on the heap; hence, it is critical that cutensornetDestroyTensorSVDInfo() is called once
svdInfo
is no longer required.- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
svdInfo – [out] This data structure holds all information about the trucation at runtime.
cutensornetDestroyTensorSVDInfo
¶
-
cutensornetStatus_t cutensornetDestroyTensorSVDInfo(cutensornetTensorSVDInfo_t svdInfo)¶
Frees all the memory associated with the TensorSVDInfo object.
- Parameters
svdInfo – [inout] Opaque handle to a TensorSVDInfo object.
cutensornetTensorSVDInfoGetAttribute
¶
-
cutensornetStatus_t cutensornetTensorSVDInfoGetAttribute(const cutensornetHandle_t handle, const cutensornetTensorSVDInfo_t svdInfo, cutensornetTensorSVDInfoAttributes_t attr, void *buf, size_t sizeInBytes)¶
Gets attributes of
svdInfo
.- Parameters
handle – [in] Opaque handle holding cuTENSORNet’s library context.
svdInfo – [in] Opaque structure that is accessed.
attr – [in] Specifies the attribute that is requested.
buf – [out] On return, this buffer (of size
sizeInBytes
) holds the value that corresponds toattr
withinsvdConfig
.sizeInBytes – [in] Size of
buf
(in bytes).
Distributed Parallelization API¶
cutensornetDistributedResetConfiguration
¶
-
cutensornetStatus_t cutensornetDistributedResetConfiguration(cutensornetHandle_t handle, const void *commPtr, size_t commSize)¶
Resets the distributed MPI parallelization configuration.
This function accepts a user-provided MPI communicator in a type-erased form and stores a copy of it inside the cuTensorNet library handle. The provided MPI communicator must be explicitly created by calling MPI_Comm_dup (please see the MPI specification). The subsequent calls to the contraction path finder, contraction plan autotuning, and contraction execution will be parallelized across all MPI processes in the provided MPI communicator. The provided MPI communicator is owned by the user, it should stay alive until the next reset call with a different MPI communicator. If NULL is provided as the pointer to the MPI communicator, no parallelization will be applied to the above mentioned procedures such that those procedures will execute redundantly across all MPI processes. As an example, please refer to the tensornet_example_mpi_auto.cu sample.
To enable distributed parallelism, cuTensorNet requires users to set an environment variable
$CUTENSORNET_COMM_LIB
containing the path to a shared library wrapping the communication primitives. For MPI users, we ship a wrapper filecutensornet_distributed_interface_mpi.c
that can be compiled against the MPI library. cuTensorNet will use the included function pointers to perform inter-process communication.Warning
This is a collective call that must be executed by all MPI processes. Note that one can still provide different (non-NULL) MPI communicators to different subgroups of MPI processes (to create concurrent cuTensorNet distributed subgroups).
The provided MPI communicator must not be used by more than one cuTensorNet library handle. This is automatically ensured by using MPI_Comm_dup.
The current library implementation assumes one GPU instance per MPI rank since the cutensornet library handle is associated with a single GPU instance. In case of multiple GPUs per node, each MPI process running on the same node may still see all GPU devices if CUDA_VISIBLE_DEVICES was not set to provide an exclusive access to each GPU. In such a case, the cutensornet library runtime will assign GPU #(processRank % numVisibleDevices), where processRank is the rank of the current process in its MPI communicator, and numVisibleDevices is the number of GPU devices visible to the current MPI process. The assigned GPU must coincide with the one associated with the cutensornet library handle, otherwise resulting in an error. To ensure consistency, the user must call cudaSetDevice in each MPI process to select the correct GPU device prior to creating a cutensornet library handle.
It is user’s responsibility to ensure that each MPI process in each provided MPI communicator executes exactly the same sequence of cutensornet API calls, which otherwise will result in an undefined behavior.
- Parameters
handle – [in] cuTensorNet library handle.
commPtr – [in] A pointer to the provided MPI communicator created by MPI_Comm_dup.
commSize – [in] The size of the provided MPI communicator: sizeof(MPI_Comm).
cutensornetDistributedGetNumRanks
¶
-
cutensornetStatus_t cutensornetDistributedGetNumRanks(const cutensornetHandle_t handle, int32_t *numRanks)¶
Queries the number of MPI ranks in the current distributed MPI configuration.
Warning
The number of ranks corresponds to the MPI communicator used by the current MPI process. If different subgroups of MPI processes used different MPI communicators, the reported number will refer to their specific MPI communicators.
- Parameters
handle – [in] cuTensorNet library handle.
numRanks – [out] Number of MPI ranks in the current distributed MPI configuration.
cutensornetDistributedGetProcRank
¶
-
cutensornetStatus_t cutensornetDistributedGetProcRank(const cutensornetHandle_t handle, int32_t *procRank)¶
Queries the rank of the current MPI process in the current distributed MPI configuration.
Warning
The MPI process rank corresponds to the MPI communicator used by that MPI process. If different subgroups of MPI processes used different MPI communicators, the reported number will refer to their specific MPI communicators.
- Parameters
handle – [in] cuTensorNet library handle.
procRank – [out] Rank of the current MPI process in the current distributed MPI configuration.
cutensornetDistributedSynchronize
¶
-
cutensornetStatus_t cutensornetDistributedSynchronize(const cutensornetHandle_t handle)¶
Globally synchronizes all MPI processes in the current distributed MPI configuration, ensuring that all preceding cutensornet API calls have completed across all MPI processes.
Warning
This is a collective call that must be executed by all MPI processes.
Prior to performing the global synchronization, the user is still required to synchronize GPU operations locally (via CUDA stream synchronization).
- Parameters
handle – [in] cuTensorNet library handle.
High-Level Tensor Network API¶
The high-level tensor network API functions allow users to define complex tensor network
states by gradually applying tensor operators to the initial (vacuum) state residing in the
user-defined tensor-product space, that is, the space constructed as a tensor product
of multiple vector spaces of given dimensions. Such a way of defining the tensor network
states is very convenient for quantum circuit simulators since the final output state
of a given quantum circuit is constructed by gradually applying quantum gates
to the initial (vacuum) state of all qubits (or qudits). Once the action of all
tensor operators have been specified, the underlying tensor network is completely defined,
together with the final state of the tensor network that resides in the same tensor-product
space as the initial (vacuum) state. At this point, different properties of the defined
tensor network state can be computed. cutensornetStateMarginal_t
and corresponding API
can be used to compute the marginal probability distribution tensors (reduced density matrices)
over the specified modes of the tensor network state. In particular, including all tensor network
state modes would result in the computation of the full density matrix of the tensor network state.
cutensornetStateSampler_t
and corresponding API can be used to sample from the probability
distribution associated with the specified tensor network state modes. For each such property
of the tensor network state that one may want to compute, the corresponding subset of the high-level
API includes API functions for defining (creating) the feature-providing object, like those named above,
configuring it by setting specific attributes, preparing it for computation, and, finally, computing it
(and, of course, destroying it at the end).
cutensornetCreateState
¶
-
cutensornetStatus_t cutensornetCreateState(const cutensornetHandle_t handle, cutensornetStatePurity_t purity, int32_t nStateModes, const int64_t *stateModeExtents, cudaDataType_t dataType, cutensornetState_t *tensorNetworkState)¶
Creates an empty tensor network state of a given shape defined by the number of primary tensor modes and their extents.
A tensor network state is the tensor representing the result of a full contraction of some (yet unspecified) tensor network. That is, a tensor network state is simply a tensor living in a given primary tensor-product space fully specified by extents of all independent dimensions of that space. A tensor network state (state tensor) can be either pure or mixed. A pure state tensor resides in the defining primary tensor-product space and is represented by the tensor belonging to that (primary) space. A mixed state tensor resides in the direct-product space formed by the defining primary tensor-product space and its dual (conjugated) space. A mixed state tensor is a tensor with twice more dimensions, namely, dimensions from the defining primary tensor-product space followed by the same number of dimensions from its dual space.
Note
To give a concrete example, a pure state tensor of any quantum circuit with 4 qubits has the shape [2,2,2,2] (quantum circuit is a specific kind of tensor network). A mixed state tensor in this case will have the shape [2,2,2,2,2,2,2,2], corresponding to the density matrix of the 4-qubit register, although there are still only 4 defining modes associated with the primary tensor-product space of 4 qubits in this case. That is, a mixed state tensor contains two sets of modes, one from the primary space and one from its dual space, but it is still defined by the modes of the primary tensor-product space.
Warning
The current cuTensorNet release only supports pure tensor network states.
- Parameters
handle – [in] cuTensorNet library handle.
purity – [in] Desired purity of the tensor network state (pure or mixed).
nStateModes – [in] Number of the defining state modes, irrespective of state purity. Note that both pure and mixed tensor network states are defined solely by the modes of the primary tensor-product space.
stateModeExents – [in] Pointer to the extents of the defining state modes.
dataType – [in] Data type of the state tensor.
tensorNetworkState – [out] Tensor network state (empty at this point).
cutensornetDestroyState
¶
-
cutensornetStatus_t cutensornetDestroyState(cutensornetState_t tensorNetworkState)¶
Destroys a tensor network state.
Note
After the tensor network state is destroyed, all pointers to the elements of tensor operators forming the state may be invalidated.
- Parameters
tensorNetworkState – [in] Tensor network state.
cutensornetStateApplyTensor
¶
-
cutensornetStatus_t cutensornetStateApplyTensor(const cutensornetHandle_t handle, cutensornetState_t tensorNetworkState, int32_t nStateModes, const int32_t *stateModes, void *tensorData, const int64_t *tensorModeStrides, const int32_t immutable, const int32_t adjoint, const int32_t unitary, int64_t *tensorId)¶
Applies a tensor operator to the tensor network state.
A tensor operator acts on the specified subset of state tensor modes, where the number of state modes defines its rank. A tensor operator is represented by a tensor with twice more dimensions than the number of state modes it acts on, where the first half of the tensor operator dimensions is contracted with the state modes of the input state tensor whereas the second half of the tensor operator dimensions forms the output state tensor modes. Since the default tensor storage strides follow the generalized columnwise layout, the action of a rank-2 tensor operator G on a rank-2 state tensor Q0 can be expressed symbolically as: Q1(i1,i0) = Q0(j1,j0) * G(j1,j0,i1,i0), which is simply the reverse form of the standard notation: Q1(i0,i1) = G(i0,i1,j0,j1) * Q0(j0,j1). In this way, one can assume the standard row-following initialization of the tensor operator.
Note
For the purpose of quantum circuit definition, our current convention conveniently allows initialization of a 2-qubit CNOT gate (tensor operator) with a C array with elements precisely following the canonical textbook (row-following) definition of the CNOT gate:
\[\begin{split} \begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 0 & 1 \\ 0 & 0 & 1 & 0 \end{pmatrix} \end{split}\]Warning
The pointer to the tensor operator elements is owned by the user and it must stay valid for the whole lifetime of the tensor network state, unless explicitly replaced by another pointer via the cutensornetStateUpdateTensor() call.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkState – [in] Tensor network state.
nStateModes – [in] Number of state modes the tensor operator acts on.
stateModes – [in] Pointer to the state modes the tensor operator acts on.
tensorData – [in] Elements of the tensor operator (must be of the same type as the state tensor).
tensorModeStrides – [in] Strides of the tensor operator data. NULL means the default generalized columnwise layout.
immutable – [in] Whether or not the tensor operator data may change during the lifetime of the tensor network state.
adjoint – [in] Whether or not the tensor operator is applied as an adjoint.
unitary – [in] Whether or not the tensor operator is unitary with respect to the first and second halves of its dimensions.
tensorId – [out] Unique integer tensor id.
cutensornetStateUpdateTensor
¶
-
cutensornetStatus_t cutensornetStateUpdateTensor(const cutensornetHandle_t handle, cutensornetState_t tensorNetworkState, int64_t tensorId, void *tensorData, int32_t unitary)¶
Updates the elements of the specified tensor operator previously applied to the tensor network state.
Warning
The pointer to the tensor operator elements is owned by the user and it must stay valid for the whole lifetime of the tensor network state, unless explicitly replaced by another pointer via the cutensornetStateUpdateTensor() call.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkState – [in] Tensor network state.
tensorId – [in] Tensor id assigned during the cutensornetStateApplyTensor() call.
tensorData – [in] Elements of the tensor operator (must be of the same type as the state tensor).
unitary – [in] Whether or not the tensor operator is unitary with respect to the first and second halves of its dimensions.
cutensornetCreateMarginal
¶
-
cutensornetStatus_t cutensornetCreateMarginal(const cutensornetHandle_t handle, cutensornetState_t tensorNetworkState, int32_t nMarginalModes, const int32_t *marginalModes, int32_t nProjectedModes, const int32_t *projectedModes, const int64_t *marginalTensorStrides, cutensornetStateMarginal_t *tensorNetworkMarginal)¶
Creates a representation of the specified marginal tensor for a given tensor network state.
The tensor network state marginal tensor is formed by a direct product of the tensor network state with its dual (conjugated) state, followed by a trace over all state modes except those explicitly specified (so-called open state modes). The order of the specified open state modes will be respected when computing the tensor network state marginal tensor. Additionally, prior to tracing, some of the state modes can optionally be projected to specific individual states of those modes, thus forming the so-called projected modes which will not be involved in tracing.
Note
In the quantum domain, the marginal tensor is known as the reduced density matrix. For example, in quantum circuit simulations, the reduced density matrix is specified by the state modes which are kept intact and the remaining state modes which are traced over. Additionally, prior to tracing, one can project certain qudit modes to specific individual states of those modes, resulting in a projected reduced density matrix.
Warning
The tensor network state must stay valid and structurally unchanged during the lifetime of the tensor network state marginal. By “structurally unchanged”, we mean no additional tensor operators can be applied to it without invalidating the currently created tensor network state marginal.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkState – [in] Tensor network state.
nMarginalModes – [in] Number of open state modes defining the marginal tensor.
marginalModes – [in] Pointer to the open state modes defining the marginal tensor.
nProjectedModes – [in] Number of projected state modes.
projectedModes – [in] Pointer to the projected state modes.
marginalTensorStrides – [in] Storage strides for the marginal tensor.
tensorNetworkMarginal – [out] Tensor network state marginal.
cutensornetDestroyMarginal
¶
-
cutensornetStatus_t cutensornetDestroyMarginal(cutensornetStateMarginal_t tensorNetworkMarginal)¶
Destroys the tensor network state marginal.
- Parameters
tensorNetworkMarginal – [in] Tensor network state marginal.
cutensornetMarginalConfigure
¶
-
cutensornetStatus_t cutensornetMarginalConfigure(const cutensornetHandle_t handle, cutensornetStateMarginal_t tensorNetworkMarginal, cutensornetMarginalAttributes_t attribute, const void *attributeValue, size_t attributeSize)¶
Configures computation of the tensor network state marginal tensor.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkMarginal – [in] Tensor network state marginal.
attribute – [in] Configuration attribute.
attributeValue – [in] Pointer to the configuration attribute value (type-erased).
attributeSize – [in] The size of the configuration attribute value.
cutensornetMarginalPrepare
¶
-
cutensornetStatus_t cutensornetMarginalPrepare(const cutensornetHandle_t handle, cutensornetStateMarginal_t tensorNetworkMarginal, size_t maxWorkspaceSizeDevice, cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t cudaStream)¶
Prepares computation of the tensor network state marginal tensor.
Warning
The cudaStream argument is unused in the current release (can be set to 0x0).
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkMarginal – [in] Tensor network state marginal.
maxWorkspaceSizeDevice – [in] Upper limit on the amount of memory (bytes) available on the GPU device for the computation of the tensor network state marginal tensor.
workDesc – [out] Workspace descriptor (the required scratch/cache memory sizes will be set).
cudaStream – [in] CUDA stream.
cutensornetMarginalCompute
¶
-
cutensornetStatus_t cutensornetMarginalCompute(const cutensornetHandle_t handle, cutensornetStateMarginal_t tensorNetworkMarginal, const int64_t *projectedModeValues, cutensornetWorkspaceDescriptor_t workDesc, void *marginalTensor, cudaStream_t cudaStream)¶
Computes the tensor network state marginal tensor.
Warning
The provided workspace descriptor
workDesc
must have the Device Scratch buffer set explicitly since user-provided memory pools are not supported in the current release. Additionally, the attached workspace buffer must be 256-byte aligned in the current release.Warning
In the current release, the execution of this API function will synchronize the provided CUDA stream. This restriction may be released in the future.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkMarginal – [in] Tensor network state marginal.
projectedModeValues – [in] Pointer to the values of the projected modes. Each integer value corresponds to a basis state of the given (projected) state mode.
workDesc – [in] Workspace descriptor (the required scratch/cache memory buffers must be set by the user).
marginalTensor – [out] Pointer to the GPU storage of the marginal tensor which will be computed in this call.
cudaStream – [in] CUDA stream.
cutensornetCreateSampler
¶
-
cutensornetStatus_t cutensornetCreateSampler(const cutensornetHandle_t handle, cutensornetState_t tensorNetworkState, int32_t nModesToSample, const int32_t *modesToSample, cutensornetStateSampler_t *tensorNetworkSampler)¶
Creates a tensor network state sampler.
A tensor network state sampler produces samples of the state tensor with the probability equal to the squared absolute value of the corresponding elements of the state tensor. One can also choose any subset of state tensor modes to sample only from the space spanned by them. The order of the specified state tensor modes will be respected when producing the output samples.
Note
For the purpose of quantum circuit simulations, the tensor network state sampler can generate bit-strings (or qudit-strings) from the output state of the defined quantum circuit (tensor network defined by gate applications).
Warning
The tensor network state must stay valid and structurally unchanged during the lifetime of the tensor network state sampler. By “structurally unchanged”, we mean no additional tensor operators can be applied to it without invalidating the currently created tensor network state sampler.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkState – [in] Tensor network state.
nModesToSample – [in] Number of state tensor modes to sample.
modesToSample – [in] Pointer to the state tensor modes to sample.
tensorNetworkSampler – [out] Tensor network sampler.
cutensornetDestroySampler
¶
-
cutensornetStatus_t cutensornetDestroySampler(cutensornetStateSampler_t tensorNetworkSampler)¶
Destroys the tensor network state sampler.
- Parameters
tensorNetworkSampler – [in] Tensor network state sampler.
cutensornetSamplerConfigure
¶
-
cutensornetStatus_t cutensornetSamplerConfigure(const cutensornetHandle_t handle, cutensornetStateSampler_t tensorNetworkSampler, cutensornetSamplerAttributes_t attribute, const void *attributeValue, size_t attributeSize)¶
Configures the tensor network state sampler.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkSampler – [in] Tensor network state sampler.
attribute – [in] Configuration attribute.
attributeValue – [in] Pointer to the configuration attribute value (type-erased).
attributeSize – [in] The size of the configuration attribute value.
cutensornetSamplerPrepare
¶
-
cutensornetStatus_t cutensornetSamplerPrepare(const cutensornetHandle_t handle, cutensornetStateSampler_t tensorNetworkSampler, size_t maxWorkspaceSizeDevice, cutensornetWorkspaceDescriptor_t workDesc, cudaStream_t cudaStream)¶
Prepares the tensor network state sampler.
Warning
The cudaStream argument is unused in the current release (can be set to 0x0).
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkSampler – [in] Tensor network state sampler.
maxWorkspaceSizeDevice – [in] Upper limit on the amount of memory (bytes) available on the GPU device for the tensor network state sampler.
workDesc – [out] Workspace descriptor (the required scratch/cache memory sizes will be set).
cudaStream – [in] CUDA stream.
cutensornetSamplerSample
¶
-
cutensornetStatus_t cutensornetSamplerSample(const cutensornetHandle_t handle, cutensornetStateSampler_t tensorNetworkSampler, int64_t nShots, cutensornetWorkspaceDescriptor_t workDesc, int64_t *samples, cudaStream_t cudaStream)¶
Performs sampling of the tensor network state, that is, generates the requested number of samples.
Note
The pseudo-random number generator used internally is initialized with a random default seed, thus generally resulting in a different set of samples generated upon each repeated execution. In future, the ability to reset the seed to a user-defined value may be provided, to ensure generation of exactly the same set of samples upon rerunning the application repeatedly (this could be useful for debugging).
Warning
The provided workspace descriptor
workDesc
must have the Device Scratch buffer set explicitly since user-provided memory pools are not supported in the current release. Additionally, the attached workspace buffer must be 256-byte aligned in the current release.Warning
In the current release, the execution of this API function will synchronize the provided CUDA stream. This restriction may be released in the future.
- Parameters
handle – [in] cuTensorNet library handle.
tensorNetworkSampler – [in] Tensor network state sampler.
nShots – [in] Number of samples to generate.
workDesc – [in] Workspace descriptor (the required scratch/cache memory buffers must be set by the user).
samples – [out] Host memory pointer to where the generated state tensor samples will be stored. The samples will be stored as samples[SampleId][ModeId] in C notation and the originally specified order of the state tensor modes to sample from will be respected.
cudaStream – [in] CUDA stream.
Memory Management API¶
A stream-ordered memory allocator (or mempool for short) allocates/deallocates memory asynchronously from/to a mempool in a
stream-ordered fashion, meaning memory operations and computations enqueued on the streams have a well-defined inter- and intra-
stream dependency. There are several well-implemented stream-ordered mempools available, such as cudaMemPool_t
that is built-in
at the CUDA driver level since CUDA 11.2 (so that all CUDA applications in the same process can easily share the same pool,
see here) and the RAPIDS
Memory Manager (RMM). For a detailed introduction, see the NVIDIA Developer Blog.
The new device memory handler APIs allow users to bind a stream-ordered mempool to the library handle, such that cuTensorNet can take care of most of the memory management for users. Below is an illustration of what can be done:
MyMemPool pool = MyMemPool(); // kept alive for the entire process in real apps
int my_alloc(void* ctx, void** ptr, size_t size, cudaStream_t stream) {
// assuming this is the memory allocation routine provided by my mempool
return reinterpret_cast<MyMemPool*>(ctx)->alloc(ptr, size, stream);
}
int my_dealloc(void* ctx, void* ptr, size_t size, cudaStream_t stream) {
// assuming this is the memory deallocation routine provided by my mempool
return reinterpret_cast<MyMemPool*>(ctx)->dealloc(ptr, size, stream);
}
// create a mem handler and fill in the required members for the library to use
cutensornetDeviceMemHandler_t handler;
handler.ctx = reinterpret_cast<void*>(&pool);
handler.device_alloc = my_alloc;
handler.device_free = my_dealloc;
memcpy(handler.name, std::string("my pool").c_str(), CUTENSORNET_ALLOCATOR_NAME_LEN);
// bind the handler to the library handle
cutensornetSetDeviceMemHandler(handle, &handler);
/* ... perform the network creation & optimization as usual ... */
// create a workspace descriptor
cutensornetWorkspaceDescriptor_t workDesc;
// (this step is optional and workDesc can be set to NULL if one just wants
// to use the "recommended" workspace size)
cutensornetCreateWorkspaceDescriptor(handle, &workDesc);
// User doesn’t compute the required sizes
// User doesn’t query the workspace size (but one can if desired)
// User doesn’t allocate memory!
// User sets workspacePtr=NULL for the corresponding memory space (device, in this case) to indicate the library should
// draw memory (of the "recommended" size, if the workspace size is set to 0 as shown below) from the user's pool;
// if a nonzero size is set, we would use the given size instead of the recommended one.
// (this step is also optional if workDesc has been set to NULL)
cutensornetWorkspaceSetMemory(handle, workDesc, CUTENSORNET_MEMSPACE_DEVICE, CUTENSORNET_WORKSPACE_SCRATCH, NULL, 0);
// create a contraction plan
cutensornetContractionPlan_t plan;
cutensornetCreateContractionPlan(handle, descNet, optimizerInfo, workDesc, &plan);
// autotune the plan with the workspace
cutensornetContractionAutotune(handle, plan, rawDataIn, rawDataOut, workDesc, pref, stream);
// perform actual contraction with the workspace
for (int sliceId=0; sliceId<num_slices; sliceId++) {
cutensornetContraction(
handle, plan, rawDataIn, rawDataOut, workDesc, sliceId, stream);
}
// clean up
cutensornetDestroyContractionPlan(plan);
cutensornetDestroyWorkspaceDescriptor(workDesc); // optional if workDesc has been set to NULL
// User doesn’t deallocate memory!
As shown above, several calls to the workspace-related APIs can be skipped. Moreover, allowing the library to share your memory pool not only can alleviate potential memory conflicts, but also enable possible optimizations.
Note
In the current release, only a device mempool can be bound.
cutensornetSetDeviceMemHandler
¶
-
cutensornetStatus_t cutensornetSetDeviceMemHandler(cutensornetHandle_t handle, const cutensornetDeviceMemHandler_t *devMemHandler)¶
Set the current device memory handler.
Once set, when cuTensorNet needs device memory in various API calls it will allocate from the user-provided memory pool and deallocate at completion. See cutensornetDeviceMemHandler_t and APIs that require cutensornetWorkspaceDescriptor_t for further detail.
The internal stream order is established using the user-provided stream passed to cutensornetContractionAutotune() and cutensornetContraction().
Warning
It is undefined behavior for the following scenarios:
the library handle is bound to a memory handler and subsequently to another handler
the library handle outlives the attached memory pool
the memory pool is not stream-ordered
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
devMemHandler – [in] the device memory handler that encapsulates the user’s mempool. The struct content is copied internally.
cutensornetGetDeviceMemHandler
¶
-
cutensornetStatus_t cutensornetGetDeviceMemHandler(const cutensornetHandle_t handle, cutensornetDeviceMemHandler_t *devMemHandler)¶
Get the current device memory handler.
- Parameters
handle – [in] Opaque handle holding cuTensorNet’s library context.
devMemHandler – [out] If previously set, the struct pointed to by
handler
is filled in, otherwise CUTENSORNET_STATUS_NO_DEVICE_ALLOCATOR is returned.
Error Management API¶
cutensornetGetErrorString
¶
-
const char *cutensornetGetErrorString(cutensornetStatus_t error)¶
Returns the description string for an error code.
Remark
non-blocking, no reentrant, and thread-safe
- Parameters
error – [in] Error code to convert to string.
- Returns
the error string
Logger API¶
cutensornetLoggerSetCallback
¶
-
cutensornetStatus_t cutensornetLoggerSetCallback(cutensornetLoggerCallback_t callback)¶
This function sets the logging callback routine.
- Parameters
callback – [in] Pointer to a callback function. Check cutensornetLoggerCallback_t.
cutensornetLoggerSetCallbackData
¶
-
cutensornetStatus_t cutensornetLoggerSetCallbackData(cutensornetLoggerCallbackData_t callback, void *userData)¶
This function sets the logging callback routine, along with user data.
- Parameters
callback – [in] Pointer to a callback function. Check cutensornetLoggerCallbackData_t.
userData – [in] Pointer to user-provided data to be used by the callback.
cutensornetLoggerSetFile
¶
-
cutensornetStatus_t cutensornetLoggerSetFile(FILE *file)¶
This function sets the logging output file.
- Parameters
file – [in] An open file with write permission.
cutensornetLoggerOpenFile
¶
-
cutensornetStatus_t cutensornetLoggerOpenFile(const char *logFile)¶
This function opens a logging output file in the given path.
- Parameters
logFile – [in] Path to the logging output file.
cutensornetLoggerSetLevel
¶
-
cutensornetStatus_t cutensornetLoggerSetLevel(int32_t level)¶
This function sets the value of the logging level.
- Parameters
level – [in] Log level, should be one of the following:
Level
Summary
Long Description
“0”
Off
logging is disabled (default)
“1”
Errors
only errors will be logged
“2”
Performance Trace
API calls that launch CUDA kernels will log their parameters and important information
“3”
Performance Hints
hints that can potentially improve the application’s performance
“4”
Heuristics Trace
provides general information about the library execution, may contain details about heuristic status
“5”
API Trace
API Trace - API calls will log their parameter and important information
cutensornetLoggerSetMask
¶
-
cutensornetStatus_t cutensornetLoggerSetMask(int32_t mask)¶
This function sets the value of the log mask.
- Parameters
mask – [in] Value of the logging mask. Masks are defined as a combination (bitwise OR) of the following masks:
Level
Description
“0”
Off
“1”
Errors
“2”
Performance Trace
“4”
Performance Hints
“8”
Heuristics Trace
“16”
API Trace
cutensornetLoggerForceDisable
¶
-
cutensornetStatus_t cutensornetLoggerForceDisable()¶
This function disables logging for the entire run.
Versioning API¶
cutensornetGetVersion
¶
-
size_t cutensornetGetVersion()¶
Returns Version number of the cuTensorNet library.
cutensornetGetCudartVersion
¶
-
size_t cutensornetGetCudartVersion()¶
Returns version number of the CUDA runtime that cuTensorNet was compiled against.
Can be compared against the CUDA runtime version from cudaRuntimeGetVersion().