OrtCUDAProviderOptions Struct Reference

CUDA Provider Options. More...

#include <onnxruntime_c_api.h>

Public Member Functions
	OrtCUDAProviderOptions ()

Public Attributes
int	device_id
	CUDA device Id Defaults to 0.
OrtCudnnConvAlgoSearch	cudnn_conv_algo_search
	CUDA Convolution algorithm search configuration. See enum OrtCudnnConvAlgoSearch for more details. Defaults to OrtCudnnConvAlgoSearchExhaustive.
size_t	gpu_mem_limit
	CUDA memory limit (To use all possible memory pass in maximum size_t) Defaults to SIZE_MAX.
int	arena_extend_strategy
	Strategy used to grow the memory arena 0 = kNextPowerOfTwo 1 = kSameAsRequested Defaults to 0.
int	do_copy_in_default_stream
	Flag indicating if copying needs to take place on the same stream as the compute stream in the CUDA EP 0 = Use separate streams for copying and compute. 1 = Use the same stream for copying and compute. Defaults to 1. WARNING: Setting this to 0 may result in data races for some models. Please see issue #4829 for more details.
int	has_user_compute_stream
	Flag indicating if there is a user provided compute stream Defaults to 0.
void *	user_compute_stream
	User provided compute stream. If provided, please set has_user_compute_stream to 1.
OrtArenaCfg *	default_memory_arena_cfg
	CUDA memory arena configuration parameters.
int	tunable_op_enable
	Enable TunableOp for using. Set it to 1/0 to enable/disable TunableOp. Otherwise, it is disabled by default. This option can be overriden by environment variable ORT_CUDA_TUNABLE_OP_ENABLE.
int	tunable_op_tuning_enable
	Enable TunableOp for tuning. Set it to 1/0 to enable/disable TunableOp tuning. Otherwise, it is disabled by default. This option can be overriden by environment variable ORT_CUDA_TUNABLE_OP_TUNING_ENABLE.
int	tunable_op_max_tuning_duration_ms
	Max tuning duration time limit for each instance of TunableOp. Defaults to 0 to disable the limit.

Detailed Description

CUDA Provider Options.

See also

OrtApi::SessionOptionsAppendExecutionProvider_CUDA

Constructor & Destructor Documentation

OrtCUDAProviderOptions()

OrtCUDAProviderOptions::OrtCUDAProviderOptions ( )

inline

Member Data Documentation

arena_extend_strategy

int OrtCUDAProviderOptions::arena_extend_strategy

Strategy used to grow the memory arena 0 = kNextPowerOfTwo
1 = kSameAsRequested
Defaults to 0.

Note

If a OrtArenaCfg has been applied, it will override this field

cudnn_conv_algo_search

OrtCudnnConvAlgoSearch OrtCUDAProviderOptions::cudnn_conv_algo_search

CUDA Convolution algorithm search configuration. See enum OrtCudnnConvAlgoSearch for more details. Defaults to OrtCudnnConvAlgoSearchExhaustive.

default_memory_arena_cfg

OrtArenaCfg* OrtCUDAProviderOptions::default_memory_arena_cfg

CUDA memory arena configuration parameters.

device_id

int OrtCUDAProviderOptions::device_id

CUDA device Id Defaults to 0.

do_copy_in_default_stream

int OrtCUDAProviderOptions::do_copy_in_default_stream

Flag indicating if copying needs to take place on the same stream as the compute stream in the CUDA EP 0 = Use separate streams for copying and compute. 1 = Use the same stream for copying and compute. Defaults to 1. WARNING: Setting this to 0 may result in data races for some models. Please see issue #4829 for more details.

gpu_mem_limit

size_t OrtCUDAProviderOptions::gpu_mem_limit

CUDA memory limit (To use all possible memory pass in maximum size_t) Defaults to SIZE_MAX.

Note

If a OrtArenaCfg has been applied, it will override this field

has_user_compute_stream

int OrtCUDAProviderOptions::has_user_compute_stream

Flag indicating if there is a user provided compute stream Defaults to 0.

tunable_op_enable

int OrtCUDAProviderOptions::tunable_op_enable

Enable TunableOp for using. Set it to 1/0 to enable/disable TunableOp. Otherwise, it is disabled by default. This option can be overriden by environment variable ORT_CUDA_TUNABLE_OP_ENABLE.

tunable_op_max_tuning_duration_ms

int OrtCUDAProviderOptions::tunable_op_max_tuning_duration_ms

Max tuning duration time limit for each instance of TunableOp. Defaults to 0 to disable the limit.

tunable_op_tuning_enable

int OrtCUDAProviderOptions::tunable_op_tuning_enable

Enable TunableOp for tuning. Set it to 1/0 to enable/disable TunableOp tuning. Otherwise, it is disabled by default. This option can be overriden by environment variable ORT_CUDA_TUNABLE_OP_TUNING_ENABLE.

user_compute_stream

void* OrtCUDAProviderOptions::user_compute_stream

User provided compute stream. If provided, please set has_user_compute_stream to 1.