Typedefs#
- group TypeDefs
Typedefs
-
typedef void *cuestPCMResults_t#
Opaque handle representing the results returned by the cuEST PCM compute routines (
cuestPCMPotentialCompute,cuestPCMDerivativeCompute,cuestPCMRadiiDerivativeCompute). Create it withcuestResultsCreateusingCUEST_PCM_RESULTSand read it withcuestResultsQuery. SeecuestPCMResultAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestHandle_t#
Opaque handle representing the cuEST library context.
-
typedef void *cuestAOShell_t#
Opaque handle representing a single atomic orbital (AO) Gaussian shell. See
cuestAOShellAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestAOBasis_t#
Opaque handle representing an atomic orbital (AO) basis set (group of AO shells). See
cuestAOBasisAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestAOPairList_t#
Opaque handle representing a pair list of atomic orbitals. No attributes may be queried from this handle at this time.
-
typedef void *cuestECPShell_t#
Opaque handle representing a single ECP shell. See
cuestECPShellAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestECPAtom_t#
Opaque handle representing an ECP Atom (group of ECP shells). See
cuestECPAtomAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestOEIntPlan_t#
Opaque handle for a one-electron integral computation plan (e.g., overlap integrals, kinetic integrals). No attributes may be queried from this handle at this time.
-
typedef void *cuestECPIntPlan_t#
Opaque handle for an effective core potential (ECP) integral computation plan. See
cuestECPIntPlanAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestDFIntPlan_t#
Opaque handle for a density-fitted integral computation plan (for DF/RI approaches). No attributes may be queried from this handle at this time.
-
typedef void *cuestDirectIntPlan_t#
Opaque handle for a direct 4-center integral computation plan. No attributes may be queried from this handle at this time.
-
typedef void *cuestAtomGrid_t#
Opaque handle representing the radial and angular grid points centered at a single atom center. See
cuestAtomGridAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestMolecularGrid_t#
Opaque handle representing the DFT integration grid for a molecule. See
cuestMolecularGridAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestXCIntPlan_t#
Opaque handle for an XC integral computation plan (e.g. XC potential and gradients). See
cuestXCIntPlanAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestPCMIntPlan_t#
Opaque handle for a PCM integral computation plan (e.g. PCM potential and gradients). See
cuestPCMIntPlanAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestGridKIntPlan_t#
Opaque handle for a Grid-K exchange integral computation plan. See
cuestGridKIntPlanAttributes_tfor attributes that may be queried from this handle.
-
typedef void *cuestHandleParameters_t#
Opaque handle representing the parameters needed for cuEST library context creation. See
cuestHandleParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_HANDLE_PARAMETERS. Set attributes withcuestParametersConfigure().9 configurable attributes
CUEST_HANDLE_PARAMETERS_CUDASTREAMcudaStream_tCUDA stream handle. NULL value allocates a CUDA stream internally.
May be NULL.
Default:
NULLCUEST_HANDLE_PARAMETERS_CUBLAScublasHandle_tcuBLAS handle. NULL value allocates a cuBLAS handle internally.
May be NULL.
Default:
NULLCUEST_HANDLE_PARAMETERS_CUSOLVERcusolverDnHandle_tcuSolver handle. NULL value allocates a cuSolver handle internally.
May be NULL.
Default:
NULLCUEST_HANDLE_PARAMETERS_MAX_GAUSS_HERMITEuint64_tMaximum number of quadrature points.
Default:
20CUEST_HANDLE_PARAMETERS_MAX_L_SOLID_HARMONICuint64_tHighest angular momentum for solid harmonics.
Default:
10CUEST_HANDLE_PARAMETERS_MAX_RYSuint64_tMaximum number of Rys points. (0 specifies the largest available).
Default:
0CUEST_HANDLE_PARAMETERS_RYS_SCHEMEcuestHandleParametersRysScheme_tSpecifies which scheme should be used to build the Rys quadrature table.
CUEST_HANDLE_PARAMETERS_JIT_CACHE_DIRconst char*JIT kernel cache directory. Empty string uses the derived default (~/.cuest_cache/cuest-cuda<N>-v<version>/). Must be a trusted, per-user, non-world-writable path (see security notes). The directory is created at cuestCreate if it does not exist. An existing directory that is a symlink, or world-writable without the sticky bit, is rejected at cuestCreate with CUEST_STATUS_INVALID_ARGUMENT; a path that exists but is not a directory, or that cannot be created, fails cuestCreate with CUEST_STATUS_EXCEPTION; an existing directory that is not writable is accepted at cuestCreate and fails at the first JIT compilation.
NUL-terminated; the value returned by a query is allocated with malloc and must be freed by the caller with C's native free(). The allocation only occurs when the query completes successfully (CUEST_STATUS_SUCCESS); on error nothing is allocated to free.
Default: "" (derived default)
CUEST_HANDLE_PARAMETERS_JIT_COMPILE_THREADSint32_tParallel JIT precompile worker threads. Must be >= 1 (a smaller value is rejected at cuestCreate with CUEST_STATUS_INVALID_ARGUMENT); clamped to an internal upper bound of 256.
Default:
16
-
typedef void *cuestAOShellParameters_t#
Opaque handle representing the parameters needed for atomic orbital (AO) Gaussian shell creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_AOSHELL_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestAOBasisParameters_t#
Opaque handle representing the parameters needed for atomic orbital (AO) basis set (group of AO shells) creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_AOBASIS_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestAOPairListParameters_t#
Opaque handle representing the parameters needed for pair list creation.
Create with
cuestParametersCreate()usingCUEST_AOPAIRLIST_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestECPShellParameters_t#
Opaque handle representing the parameters needed for effective core potential (ECP) shell creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_ECPSHELL_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestECPAtomParameters_t#
Opaque handle representing the parameters needed for effective core potential (ECP) atom set (group of ECP shells) creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_ECPATOM_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestOEIntPlanParameters_t#
Opaque handle representing the parameters needed for a one-electron integral computation plan (e.g., overlap integrals, kinetic integrals) creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_OEINTPLAN_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestECPIntPlanParameters_t#
Opaque handle representing the parameters needed for an effective core potential (ECP) integral computation plan creation. See
cuestECPIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_ECPINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().6 configurable attributes
CUEST_ECPINTPLAN_PARAMETERS_THRESHOLD_COLLOCATIONdoubleGrid collocation threshold.
Default:
1.0e-18CUEST_ECPINTPLAN_PARAMETERS_ECP_CUTOFFdoubleGrid cutoff.
Default:
1.0e-14CUEST_ECPINTPLAN_PARAMETERS_R0doubleGrid R0.
Default:
6.0CUEST_ECPINTPLAN_PARAMETERS_MULTIPLIERdoubleGrid multiplier.
Default:
1.1CUEST_ECPINTPLAN_PARAMETERS_RADIAL_NUM_POINTuint64_tNumber radial grid points.
Default:
80CUEST_ECPINTPLAN_PARAMETERS_SPHERICAL_NUM_POINTuint64_tNumber spherical grid points.
Default:
302
-
typedef void *cuestECPComputeParameters_t#
Opaque handle representing the parameters needed for an effective core potential (ECP) integral compute. See
cuestECPComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_ECPCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_ECPCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the local potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_ECPCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the local potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestECPDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for an effective core potential (ECP) integral derivative compute. See
cuestECPDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_ECPDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_ECPDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_ECPDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestDFIntPlanParameters_t#
Opaque handle representing the parameters needed for a density-fitted integral computation plan (for DF/RI approaches) creation. See
cuestDFIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().10 configurable attributes
CUEST_DFINTPLAN_PARAMETERS_FITTING_CUTOFFdoubleValue below which metric eigenvalues are discarded during fitting.
Default:
1.0e-12CUEST_DFINTPLAN_PARAMETERS_FITTING_RELATIVE_CONDITIONINGint32_tShould the fitting cutoff be determined in a relative or absolute sense? 0 is absolute. non-zero is relative.
Default:
1CUEST_DFINTPLAN_PARAMETERS_EXCHANGE_FRACTIONdoubleFraction of HF exchange: respected in DF exchange and DF derivative compute routines.
Default:
1.0CUEST_DFINTPLAN_PARAMETERS_FITTING_ALGORITHMcuestDFIntPlanParametersFittingAlgorithm_tAlgorithm to use when inverting the DF metric.
CUEST_DFINTPLAN_PARAMETERS_LRC_EXCHANGE_FRACTIONdoubleFraction of long-range exchange: respected in DF exchange and DF derivative compute routines.
Default:
0.0CUEST_DFINTPLAN_PARAMETERS_LRC_EXCHANGE_OMEGAdoubleOmega range-separation parameter (in inverse bohr): respected in DF exchange and DF derivative compute routines. Must be greater than or equal to 0.0.
Default:
0.0CUEST_DFINTPLAN_PARAMETERS_THREE_INDEX_INTEGRAL_DIRECTint32_tSkips the formation of 3-index integrals in the plan creation and uses integral direct algorithms in subsequent compute calls (where possible).
Default:
0CUEST_DFINTPLAN_PARAMETERS_CACHE_OPERATOR_WEIGHTINGcuestDFIntPlanParametersCacheOperatorWeighting_tWhether the cached 3-index integrals carry the operator weighting. Ignored when
CUEST_DFINTPLAN_PARAMETERS_THREE_INDEX_INTEGRAL_DIRECTis non-zero.Default:
CUEST_DFINTPLAN_PARAMETERS_CACHE_OPERATOR_WEIGHTING_FULLCUEST_DFINTPLAN_PARAMETERS_DETAIL_THREE_INDEX_FITTING_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the three-index fitting GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFINTPLAN_PARAMETERS_DETAIL_THREE_INDEX_FITTING_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the three-index fitting GEMMs. Must be in [1, 256].
Default:
63
-
typedef void *cuestAtomGridParameters_t#
Opaque handle representing the parameters needed for atomic grid plan creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_ATOMGRID_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestMolecularGridParameters_t#
Opaque handle representing the parameters needed for molecular grid plan creation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_MOLECULARGRID_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestXCIntPlanParameters_t#
Opaque handle representing the parameters needed for an exchange-correlation integral computation plan creation. See
cuestXCIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().3 configurable attributes
CUEST_XCINTPLAN_PARAMETERS_THRESHOLD_COLLOCATIONdoubleXC collocation threshold.
Default:
1.0e-18CUEST_XCINTPLAN_PARAMETERS_DERIVATIVE_LEVELint32_tMaximum derivative level supported.
Default:
1CUEST_XCINTPLAN_PARAMETERS_NUM_THREADSint32_tHow many CPU threads to use when computing Vxc. Must be >= 1.
Default:
12
-
typedef void *cuestGridKIntPlanParameters_t#
Opaque handle representing the parameters needed for a Grid-K exchange integral computation plan creation. See
cuestGridKIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_GRIDKINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().7 configurable attributes
CUEST_GRIDKINTPLAN_PARAMETERS_THRESHOLD_COLLOCATIONdoubleAO collocation threshold.
Default:
1.0e-18CUEST_GRIDKINTPLAN_PARAMETERS_EXCHANGE_FRACTIONdoubleFraction of full-range (1/r) exchange.
Default:
1.0CUEST_GRIDKINTPLAN_PARAMETERS_LRC_EXCHANGE_FRACTIONdoubleFraction of long-range (erf(wr)/r) exchange.
Default:
0.0CUEST_GRIDKINTPLAN_PARAMETERS_LRC_EXCHANGE_OMEGAdoubleOmega range-separation parameter (in inverse bohr). Must be >= 0.0.
Default:
0.0CUEST_GRIDKINTPLAN_PARAMETERS_FITTING_ALGORITHMcuestGridKIntPlanParametersFittingAlgorithm_tAlgorithm used to fit the collocation of the grid-based exchange.
Default:
CUEST_GRIDKINTPLAN_PARAMETERS_FITTING_ALGORITHM_NONECUEST_GRIDKINTPLAN_PARAMETERS_FITTING_ORTHOGONALIZATION_CUTOFFdoubleAbsolute overlap eigenvalue cutoff for the canonical orthogonalization underlying the fit. Must be > 0.0.
Default:
1.0e-6CUEST_GRIDKINTPLAN_PARAMETERS_FITTING_CUTOFFdoubleAbsolute cutoff on the eigenvalues of the orthonormalized quadrature overlap (dimensionless recovered-norm fractions, unity for an exact grid); directions below the cutoff (including any with negative quadrature weight content) are excluded from the fit. Must be in (0.0, 1.0).
Default:
1.0e-2
-
typedef void *cuestGridKSymmetricExchangeComputeParameters_t#
Opaque handle representing the parameters needed for a Grid-K symmetric exchange computation. There are no configurable parameters at this time.
Create with
cuestParametersCreate()usingCUEST_GRIDKSYMMETRICEXCHANGECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestGridKSymmetricExchangeDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for a Grid-K symmetric exchange derivative (nuclear gradient) computation. There are no configurable parameters at this time.
Create with
cuestParametersCreate()usingCUEST_GRIDKSYMMETRICEXCHANGEDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestGridKNonsymmetricExchangeComputeParameters_t#
Opaque handle representing the parameters needed for a Grid-K non-symmetric exchange computation. There are no configurable parameters at this time.
Create with
cuestParametersCreate()usingCUEST_GRIDKNONSYMMETRICEXCHANGECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestOverlapComputeParameters_t#
Opaque handle representing the parameters needed for overlap matrix computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_OVERLAPCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestOverlapDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for overlap gradient computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_OVERLAPDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestKineticComputeParameters_t#
Opaque handle representing the parameters needed for kinetic matrix computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_KINETICCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestKineticDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for kinetic gradient computation.
Create with
cuestParametersCreate()usingCUEST_KINETICDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestPotentialComputeParameters_t#
Opaque handle representing the parameters needed for potential matrix computation. See
cuestPotentialComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_POTENTIALCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_POTENTIALCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_POTENTIALCOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestPotentialDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for potential gradient computation. See
cuestPotentialDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_POTENTIALDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_POTENTIALDERIVATIVECOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_OFFCUEST_POTENTIALDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestDFCoulombComputeParameters_t#
Opaque handle representing the parameters needed for density-fitted Coulomb matrix computation. See
cuestDFCoulombComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFCOULOMBCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DFCOULOMBCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_DFCOULOMBCOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestDirectCoulombComputeParameters_t#
Opaque handle representing the parameters needed for direct 4-center Coulomb (J) matrix computation. See
cuestDirectCoulombComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DIRECTCOULOMBCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DIRECTCOULOMBCOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwiseCUEST_DIRECTCOULOMBCOMPUTE_PARAMETERS_DENSITY_SCHWARZ_THRESHOLDdoubleDensity-weighted Cauchy-Schwarz screening threshold (>= 0; 0 disables screening).
Default:
1.0e-13
-
typedef void *cuestDirectSymmetricExchangeComputeParameters_t#
Opaque handle representing the parameters needed for direct 4-center symmetric exchange (K) matrix computation (the K operator may fuse a full-range and a range-separated term; there is no fused J-with-K entry point). See
cuestDirectSymmetricExchangeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DIRECTSYMMETRICEXCHANGECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DIRECTSYMMETRICEXCHANGECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwiseCUEST_DIRECTSYMMETRICEXCHANGECOMPUTE_PARAMETERS_DENSITY_SCHWARZ_THRESHOLDdoubleDensity-weighted Cauchy-Schwarz screening threshold (>= 0; 0 disables screening).
Default:
1.0e-13
-
typedef void *cuestDirectCoulombDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for direct 4-center Coulomb (J) gradient computation. See
cuestDirectCoulombDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DIRECTCOULOMBDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DIRECTCOULOMBDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwiseCUEST_DIRECTCOULOMBDERIVATIVECOMPUTE_PARAMETERS_DENSITY_SCHWARZ_THRESHOLDdoubleDensity-weighted Cauchy-Schwarz screening threshold (>= 0; 0 disables screening).
Default:
1.0e-13
-
typedef void *cuestDirectSymmetricExchangeDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for direct 4-center symmetric exchange (K) gradient computation (the K operator may fuse a full-range and a range-separated term). See
cuestDirectSymmetricExchangeDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DIRECTSYMMETRICEXCHANGEDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DIRECTSYMMETRICEXCHANGEDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwiseCUEST_DIRECTSYMMETRICEXCHANGEDERIVATIVECOMPUTE_PARAMETERS_DENSITY_SCHWARZ_THRESHOLDdoubleDensity-weighted Cauchy-Schwarz screening threshold (>= 0; 0 disables screening).
Default:
1.0e-13
-
typedef void *cuestDirectIntPlanParameters_t#
Opaque handle representing the parameters needed for direct 4-center integral plan creation. See
cuestDirectIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DIRECTINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().3 configurable attributes
CUEST_DIRECTINTPLAN_PARAMETERS_EXCHANGE_FRACTIONdoubleFraction of full-range (1/r) exchange.
Default:
1.0CUEST_DIRECTINTPLAN_PARAMETERS_LRC_EXCHANGE_FRACTIONdoubleFraction of long-range (erf(wr)/r) exchange.
Default:
0.0CUEST_DIRECTINTPLAN_PARAMETERS_LRC_EXCHANGE_OMEGAdoubleOmega range-separation parameter (in inverse bohr). Must be >= 0.0.
Default:
0.0
-
typedef void *cuestDFSymmetricExchangeComputeParameters_t#
Opaque handle representing the parameters needed for density-fitted symmetric exchange matrix computation. See
cuestDFSymmetricExchangeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFSYMMETRICEXCHANGECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DFSYMMETRICEXCHANGECOMPUTE_PARAMETERS_INT8_SLICE_COUNTuint64_tNumber of slices used in cuBLAS emulated GEMMs. Mantissa bit count = 8 * slices - 1.
Default:
5CUEST_DFSYMMETRICEXCHANGECOMPUTE_PARAMETERS_INT8_MODULUS_COUNTuint64_tNumber of moduli used in custom Ozaki-II emulated GEMMs.
Default:
8
-
typedef void *cuestDFNonsymmetricExchangeComputeParameters_t#
Opaque handle representing the parameters needed for density-fitted non-symmetric exchange matrix computation. See
cuestDFNonsymmetricExchangeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFNONSYMMETRICEXCHANGECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_DFNONSYMMETRICEXCHANGECOMPUTE_PARAMETERS_INT8_SLICE_COUNTuint64_tNumber of slices used in cuBLAS emulated GEMMs. Mantissa bit count = 8 * slices - 1.
Default:
5CUEST_DFNONSYMMETRICEXCHANGECOMPUTE_PARAMETERS_INT8_MODULUS_COUNTuint64_tNumber of moduli used in custom Ozaki-II emulated GEMMs.
Default:
8
-
typedef void *cuestDFSymmetricDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for density-fitted gradient computation. See
cuestDFSymmetricDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().13 configurable attributes
CUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_MEMORY_POLICYcuestDFSymmetricDerivativeComputeMemoryPolicy_tSpecifies how memory should be used during the derivative evaluation.
Default:
CUEST_DFSYMMETRICDERIVATIVECOMPUTE_MEMORY_POLICY_DEVICECACHECUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwiseCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_THREE_INDEX_FITTING_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the refit of the three-index tensor under
CUEST_DFSYMMETRICDERIVATIVECOMPUTE_MEMORY_POLICY_RECOMPUTE.Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_THREE_INDEX_FITTING_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the three-index refit GEMMs. Must be in [1, 256].
Default:
63CUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_MO_TRANSFORMATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the MO transformation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_MO_TRANSFORMATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the MO transformation GEMMs. Must be in [1, 256].
Default:
63CUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_EXCHANGE_DENSITY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the exchange density GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_EXCHANGE_DENSITY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the exchange density GEMMs. Must be in [1, 256].
Default:
63CUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_METRIC_CONTRACTION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the metric contraction GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_METRIC_CONTRACTION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the metric contraction GEMMs. Must be in [1, 256].
Default:
63CUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_PSEUDODENSITY_FITTING_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the pseudo-density fitting GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_DFSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_PSEUDODENSITY_FITTING_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the pseudo-density fitting GEMMs. Must be in [1, 256].
Default:
63
-
typedef void *cuestXCPotentialRKSComputeParameters_t#
Opaque handle representing the parameters needed for XC RKS potential computation. See
cuestXCPotentialRKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCPOTENTIALRKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_XCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_XCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCPotentialUKSComputeParameters_t#
Opaque handle representing the parameters needed for XC UKS potential computation. See
cuestXCPotentialUKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCPOTENTIALUKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_XCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_XCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCDerivativeRKSComputeParameters_t#
Opaque handle representing the parameters needed for XC RKS gradient computation. See
cuestXCDerivativeRKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCDERIVATIVERKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_XCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_XCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCDerivativeUKSComputeParameters_t#
Opaque handle representing the parameters needed for XC UKS gradient computation. See
cuestXCDerivativeUKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCDERIVATIVEUKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_XCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_XCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestNonlocalXCPotentialRKSComputeParameters_t#
Opaque handle representing the parameters needed for nonlocal XC RKS potential computation. See
cuestNonlocalXCPotentialRKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().8 configurable attributes
CUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_VV10_SCALEdoubleGlobal scaling for VV10 nonlocal correlation.
Default:
0.0CUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_VV10_CdoubleVV10 C parameter.
Default:
0.0CUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_VV10_BdoubleVV10 b parameter.
Default:
0.0CUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALRKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestNonlocalXCPotentialUKSComputeParameters_t#
Opaque handle representing the parameters needed for nonlocal XC UKS potential computation. See
cuestNonlocalXCPotentialUKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().8 configurable attributes
CUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_VV10_SCALEdoubleGlobal scaling for VV10 nonlocal correlation.
Default:
0.0CUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_VV10_CdoubleVV10 C parameter.
Default:
0.0CUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_VV10_BdoubleVV10 b parameter.
Default:
0.0CUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCPOTENTIALUKSCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestNonlocalXCDerivativeRKSComputeParameters_t#
Opaque handle representing the parameters needed for nonlocal XC RKS gradient computation. See
cuestNonlocalXCDerivativeRKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().8 configurable attributes
CUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_VV10_SCALEdoubleGlobal scaling for VV10 nonlocal correlation.
Default:
0.0CUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_VV10_CdoubleVV10 C parameter.
Default:
0.0CUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_VV10_BdoubleVV10 b parameter.
Default:
0.0CUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVERKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestNonlocalXCDerivativeUKSComputeParameters_t#
Opaque handle representing the parameters needed for nonlocal XC UKS gradient computation. See
cuestNonlocalXCDerivativeUKSComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().8 configurable attributes
CUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_VV10_SCALEdoubleGlobal scaling for VV10 nonlocal correlation.
Default:
0.0CUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_VV10_CdoubleVV10 C parameter.
Default:
0.0CUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_VV10_BdoubleVV10 b parameter.
Default:
0.0CUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55CUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_NONLOCALXCDERIVATIVEUKSCOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCDensityComputeParameters_t#
Opaque handle representing the parameters needed for XC density computation. See
cuestXCDensityComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCDENSITYCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_XCDENSITYCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDENSITYCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCNonsymmetricDensityComputeParameters_t#
Opaque handle representing the parameters needed for XC nonsymmetric density computation. See
cuestXCNonsymmetricDensityComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCNONSYMMETRICDENSITYCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_XCNONSYMMETRICDENSITYCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the density collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCNONSYMMETRICDENSITYCOMPUTE_PARAMETERS_DETAIL_DENSITY_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the density collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCPotentialComputeParameters_t#
Opaque handle representing the parameters needed for XC potential computation. See
cuestXCPotentialComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCPOTENTIALCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_XCPOTENTIALCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the potential assembly GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCPOTENTIALCOMPUTE_PARAMETERS_DETAIL_POTENTIAL_ASSEMBLY_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the potential assembly GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for XC derivative computation. See
cuestXCDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_XCDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCNonsymmetricDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for XC nonsymmetric derivative computation. See
cuestXCNonsymmetricDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_XCNONSYMMETRICDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_XCNONSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_USAGE_MODEcuestEmulationUsageMode_tControls fixed point emulation in the gradient collocation GEMMs.
Default:
CUEST_EMULATION_USAGE_MODE_DEFAULTCUEST_XCNONSYMMETRICDERIVATIVECOMPUTE_PARAMETERS_DETAIL_GRADIENT_COLLOCATION_FIXEDPOINT_EMULATION_MANTISSA_BIT_COUNTint32_tMantissa bits for fixed point emulation in the gradient collocation GEMMs. Must be in [1, 256].
Default:
55
-
typedef void *cuestXCGridDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for XC grid derivative computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_XCGRIDDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestXCIntegrationGridComputeParameters_t#
Opaque handle representing the parameters needed for XC integration grid computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_XCINTEGRATIONGRIDCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestXCIntegrationWeightComputeParameters_t#
Opaque handle representing the parameters needed for XC integration weight computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_XCINTEGRATIONWEIGHTCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestPCMIntPlanParameters_t#
Opaque handle representing the parameters needed for a polarizable continuum model (PCM) integral computation plan creation. See
cuestPCMIntPlanParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_PCMINTPLAN_PARAMETERS. Set attributes withcuestParametersConfigure().2 configurable attributes
CUEST_PCMINTPLAN_PARAMETERS_CUTOFFdoubleCutoff for surface points' iSWIG switching function, below which the point is considered inactive.
Default:
1e-12CUEST_PCMINTPLAN_PARAMETERS_X_PREFACTORdoublePCM prefactor $ f = (\epsilon - 1)/(\epsilon + x) $ , where $ x==0.0 $ gives CPCM and $ x==1/2$ corresponds to COSMO.
Default:
0.0
-
typedef void *cuestPCMPotentialComputeParameters_t#
Opaque handle representing the parameters needed for PCM potential computation. See
cuestPCMPotentialComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_PCMPOTENTIALCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_PCMPOTENTIALCOMPUTE_PARAMETERS_CONVERGENCE_THRESHOLDdoubleConvergence threshold (max absolute residual value) for PCM charges computation.
Default:
1e-10CUEST_PCMPOTENTIALCOMPUTE_PARAMETERS_MAX_ITERATIONSuint64_tMaximum number of preconditioned conjugate gradient iterations for PCM charges computation.
Default:
200CUEST_PCMPOTENTIALCOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_PCMPOTENTIALCOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestPCMDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for PCM derivative computation. See
cuestPCMDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_PCMDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_PCMDERIVATIVECOMPUTE_PARAMETERS_CONVERGENCE_THRESHOLDdoubleConvergence threshold (max absolute residual value) for PCM charges computation.
Default:
1e-10CUEST_PCMDERIVATIVECOMPUTE_PARAMETERS_MAX_ITERATIONSuint64_tMaximum number of preconditioned conjugate gradient iterations for PCM charges computation.
Default:
200CUEST_PCMDERIVATIVECOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_PCMDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestPCMRadiiDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for PCM radii derivative computation. See
cuestPCMRadiiDerivativeComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_PCMRADIIDERIVATIVECOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().4 configurable attributes
CUEST_PCMRADIIDERIVATIVECOMPUTE_PARAMETERS_CONVERGENCE_THRESHOLDdoubleConvergence threshold (max absolute residual value) for PCM charges computation.
Default:
1e-10CUEST_PCMRADIIDERIVATIVECOMPUTE_PARAMETERS_MAX_ITERATIONSuint64_tMaximum number of preconditioned conjugate gradient iterations for PCM charges computation.
Default:
200CUEST_PCMRADIIDERIVATIVECOMPUTE_PARAMETERS_FFLOAT_USAGE_MODEcuestFfloatUsageMode_tParameter to control the use of ffloat compute.
Default:
CUEST_FFLOAT_USAGE_MODE_DEFAULTCUEST_PCMRADIIDERIVATIVECOMPUTE_PARAMETERS_JIT_USAGE_MODEcuestJITUsageMode_tParameter to control the use of JIT-compiled kernels.
Default:
CUEST_JIT_USAGE_MODE_ONon CUDA 13.x+,CUEST_JIT_USAGE_MODE_OFFotherwise
-
typedef void *cuestMultipoleComputeParameters_t#
Opaque handle representing the parameters needed for multipole matrix computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_MULTIPOLECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestMultipoleDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for multipole gradient computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_MULTIPOLEDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestAngularMomentumComputeParameters_t#
Opaque handle representing the parameters needed for angular momentum integral matrix computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_ANGULARMOMENTUMCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestAngularMomentumDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for angular momentum gradient computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_ANGULARMOMENTUMDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestNablaComputeParameters_t#
Opaque handle representing the parameters needed for nabla integral matrix computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_NABLACOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestNablaDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for nabla gradient computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_NABLADERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestPCMIntegrationGridComputeParameters_t#
Opaque handle representing the parameters needed for PCM integration grid computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_PCMINTEGRATIONGRIDCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestPCMIntegrationWeightComputeParameters_t#
Opaque handle representing the parameters needed for PCM integration weight computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_PCMINTEGRATIONWEIGHTCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestDFMOIntegralsComputeParameters_t#
Opaque handle representing the parameters needed for MO transformation. See
cuestDFMOIntegralsComputeParametersAttributes_tfor configurable parameters and their default values.Create with
cuestParametersCreate()usingCUEST_DFMOINTEGRALSCOMPUTE_PARAMETERS. Set attributes withcuestParametersConfigure().1 configurable attribute
CUEST_DFMOINTEGRALSCOMPUTE_PARAMETERS_OPERATOR_WEIGHTINGcuestDFMOIntegralsComputeParametersOperatorWeighting_tWhether the integrals carry the operator weighting.
Default:
CUEST_DFMOINTEGRALSCOMPUTE_PARAMETERS_OPERATOR_WEIGHTING_FULL
-
typedef void *cuestDFMOIntegralsDerivativeComputeParameters_t#
Opaque handle representing the parameters needed for the MO/TPDM gradient. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_DFMOINTEGRALSDERIVATIVECOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestOverlapHessianSkeletonComputeParameters_t#
Opaque handle representing the parameters needed for overlap Hessian skeleton computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_OVERLAPHESSIANSKELETONCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestOverlapHessianRHSComputeParameters_t#
Opaque handle representing the parameters needed for overlap Hessian RHS computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_OVERLAPHESSIANRHSCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestKineticHessianSkeletonComputeParameters_t#
Opaque handle representing the parameters needed for kinetic Hessian skeleton computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_KINETICHESSIANSKELETONCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestKineticHessianRHSComputeParameters_t#
Opaque handle representing the parameters needed for kinetic Hessian RHS computation. There are no configurable parameters for this handle at this time.
Create with
cuestParametersCreate()usingCUEST_KINETICHESSIANRHSCOMPUTE_PARAMETERS. There are no configurable attributes.
-
typedef void *cuestPCMResults_t#