Kernel Tuner
1.0
Kernel Tuner
Introduction
Installation
Guides
Getting Started
Convolution
Diffusion
Matrix multiplication
Kernel Tuner Examples
Features
Backends
Cache files
Correctness Verification
Tuning Host Code
Using structs
Templated kernels
Optimization strategies
Metrics and Objectives
Observers
Reference
API Documentation
Parameter Vocabulary
Design documentation
Contribution guide
Kernel Tuner
Index
Edit on GitHub
Index
_
|
A
|
B
|
C
|
D
|
E
|
F
|
G
|
H
|
I
|
K
|
L
|
M
|
N
|
O
|
P
|
R
|
S
|
T
|
U
|
W
_
__init__() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
(kernel_tuner.runners.sequential.SequentialRunner method)
(kernel_tuner.runners.simulation.SimulationRunner method)
A
acceptance_prob() (in module kernel_tuner.strategies.simulated_annealing)
after_finish() (kernel_tuner.observers.BenchmarkObserver method)
after_start() (kernel_tuner.observers.BenchmarkObserver method)
B
before_start() (kernel_tuner.observers.BenchmarkObserver method)
benchmark() (kernel_tuner.core.DeviceInterface method)
benchmark_continuous() (kernel_tuner.core.DeviceInterface method)
benchmark_default() (kernel_tuner.core.DeviceInterface method)
BenchmarkObserver (class in kernel_tuner.observers)
C
check_argument_list() (in module kernel_tuner.util)
check_argument_type() (in module kernel_tuner.util)
check_kernel_output() (kernel_tuner.core.DeviceInterface method)
check_restrictions() (in module kernel_tuner.util)
check_stop_criterion() (in module kernel_tuner.util)
check_thread_block_dimensions() (in module kernel_tuner.util)
check_tune_params_list() (in module kernel_tuner.util)
cleanup_lib() (kernel_tuner.backends.compiler.CompilerFunctions method)
CompilationFailedConfig (class in kernel_tuner.util)
compile() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
compile_kernel() (kernel_tuner.core.DeviceInterface method)
compile_restrictions() (in module kernel_tuner.util)
CompilerFunctions (class in kernel_tuner.backends.compiler)
compute_intensity() (kernel_tuner.strategies.firefly_algorithm.Firefly method)
config_valid() (in module kernel_tuner.util)
convert_constraint_restriction() (in module kernel_tuner.util)
copy_constant_memory_args() (kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
copy_shared_memory_args() (kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
copy_texture_memory_args() (kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
correct_open_cache() (in module kernel_tuner.util)
create_device_targets() (in module kernel_tuner)
create_kernel_instance() (kernel_tuner.core.DeviceInterface method)
cuda_error_check() (in module kernel_tuner.util)
CudaFunctions (class in kernel_tuner.backends.nvcuda)
CupyFunctions (class in kernel_tuner.backends.cupy)
D
default() (kernel_tuner.util.NpEncoder method)
delete_temp_file() (in module kernel_tuner.util)
detect_language() (in module kernel_tuner.util)
DeviceInterface (class in kernel_tuner.core)
disruptive_uniform_crossover() (in module kernel_tuner.strategies.genetic_algorithm)
distance_to() (kernel_tuner.strategies.firefly_algorithm.Firefly method)
dump_cache() (in module kernel_tuner.util)
during() (kernel_tuner.observers.BenchmarkObserver method)
E
ErrorConfig (class in kernel_tuner.util)
F
Firefly (class in kernel_tuner.strategies.firefly_algorithm)
G
generate_normalized_param_dicts() (in module kernel_tuner.strategies.bayes_opt)
get_best_config() (in module kernel_tuner.util)
get_config_string() (in module kernel_tuner.util)
get_environment() (kernel_tuner.core.DeviceInterface method)
get_grid_dimensions() (in module kernel_tuner.util)
get_instance_string() (in module kernel_tuner.util)
get_kernel_string() (in module kernel_tuner.util)
get_options() (in module kernel_tuner.strategies.common)
get_problem_size() (in module kernel_tuner.util)
get_results() (kernel_tuner.observers.BenchmarkObserver method)
get_smem_args() (in module kernel_tuner.util)
get_strategy_docstring() (in module kernel_tuner.strategies.common)
get_temp_filename() (in module kernel_tuner.util)
get_thread_block_dimensions() (in module kernel_tuner.util)
get_total_timings() (in module kernel_tuner.util)
H
HipFunctions (class in kernel_tuner.backends.hip)
I
InvalidConfig (class in kernel_tuner.util)
K
kernel_finished() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
kernel_tuner.strategies.basinhopping
module
kernel_tuner.strategies.bayes_opt
module
kernel_tuner.strategies.brute_force
module
kernel_tuner.strategies.common
module
kernel_tuner.strategies.diff_evo
module
kernel_tuner.strategies.dual_annealing
module
kernel_tuner.strategies.firefly_algorithm
module
kernel_tuner.strategies.genetic_algorithm
module
kernel_tuner.strategies.greedy_ils
module
kernel_tuner.strategies.greedy_mls
module
kernel_tuner.strategies.minimize
module
kernel_tuner.strategies.mls
module
kernel_tuner.strategies.ordered_greedy_mls
module
kernel_tuner.strategies.pso
module
kernel_tuner.strategies.random_sample
module
kernel_tuner.strategies.simulated_annealing
module
kernel_tuner.util
module
L
looks_like_a_filename() (in module kernel_tuner.util)
M
make_strategy_options_doc() (in module kernel_tuner.strategies.common)
memcpy_dtoh() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions static method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
memcpy_htod() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions static method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
memset() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions static method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
module
kernel_tuner.strategies.basinhopping
kernel_tuner.strategies.bayes_opt
kernel_tuner.strategies.brute_force
kernel_tuner.strategies.common
kernel_tuner.strategies.diff_evo
kernel_tuner.strategies.dual_annealing
kernel_tuner.strategies.firefly_algorithm
kernel_tuner.strategies.genetic_algorithm
kernel_tuner.strategies.greedy_ils
kernel_tuner.strategies.greedy_mls
kernel_tuner.strategies.minimize
kernel_tuner.strategies.mls
kernel_tuner.strategies.ordered_greedy_mls
kernel_tuner.strategies.pso
kernel_tuner.strategies.random_sample
kernel_tuner.strategies.simulated_annealing
kernel_tuner.util
move_towards() (kernel_tuner.strategies.firefly_algorithm.Firefly method)
mutate() (in module kernel_tuner.strategies.genetic_algorithm)
N
neighbor() (in module kernel_tuner.strategies.simulated_annealing)
normalize_parameter_space() (in module kernel_tuner.strategies.bayes_opt)
normalize_verify_function() (in module kernel_tuner.util)
NpEncoder (class in kernel_tuner.util)
NVMLObserver (class in kernel_tuner.observers.nvml)
O
OpenCLFunctions (class in kernel_tuner.backends.opencl)
P
parse_restrictions() (in module kernel_tuner.util)
PMTObserver (class in kernel_tuner.observers.pmt)
PowerSensorObserver (class in kernel_tuner.observers.powersensor)
prepare_kernel_string() (in module kernel_tuner.util)
preprocess_gpu_arguments() (kernel_tuner.core.DeviceInterface static method)
print_config() (in module kernel_tuner.util)
print_config_output() (in module kernel_tuner.util)
process_cache() (in module kernel_tuner.util)
process_metrics() (in module kernel_tuner.util)
prune_parameter_space() (in module kernel_tuner.strategies.bayes_opt)
PyCudaFunctions (class in kernel_tuner.backends.pycuda)
R
read_cache() (in module kernel_tuner.util)
read_file() (in module kernel_tuner.util)
ready_argument_list() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
register_configuration() (kernel_tuner.observers.BenchmarkObserver method)
register_device() (kernel_tuner.observers.BenchmarkObserver method)
replace_param_occurrences() (in module kernel_tuner.util)
run() (kernel_tuner.runners.sequential.SequentialRunner method)
(kernel_tuner.runners.simulation.SimulationRunner method)
run_kernel() (in module kernel_tuner)
(kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
(kernel_tuner.core.DeviceInterface method)
RuntimeFailedConfig (class in kernel_tuner.util)
S
scale_from_params() (in module kernel_tuner.strategies.common)
SequentialRunner (class in kernel_tuner.runners.sequential)
setup_block_and_grid() (in module kernel_tuner.util)
setup_method_arguments() (in module kernel_tuner.strategies.common)
setup_method_options() (in module kernel_tuner.strategies.common)
SimulationRunner (class in kernel_tuner.runners.simulation)
single_point_crossover() (in module kernel_tuner.strategies.genetic_algorithm)
SkippableFailure
snap_to_nearest_config() (in module kernel_tuner.strategies.common)
start_event() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
stop_event() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
StopCriterionReached
store_cache() (in module kernel_tuner.util)
store_results() (in module kernel_tuner)
synchronize() (kernel_tuner.backends.compiler.CompilerFunctions method)
(kernel_tuner.backends.cupy.CupyFunctions method)
(kernel_tuner.backends.hip.HipFunctions method)
(kernel_tuner.backends.nvcuda.CudaFunctions static method)
(kernel_tuner.backends.opencl.OpenCLFunctions method)
(kernel_tuner.backends.pycuda.PyCudaFunctions method)
T
tune() (in module kernel_tuner.strategies.basinhopping)
(in module kernel_tuner.strategies.bayes_opt)
(in module kernel_tuner.strategies.brute_force)
(in module kernel_tuner.strategies.diff_evo)
(in module kernel_tuner.strategies.dual_annealing)
(in module kernel_tuner.strategies.firefly_algorithm)
(in module kernel_tuner.strategies.genetic_algorithm)
(in module kernel_tuner.strategies.greedy_ils)
(in module kernel_tuner.strategies.greedy_mls)
(in module kernel_tuner.strategies.minimize)
(in module kernel_tuner.strategies.mls)
(in module kernel_tuner.strategies.ordered_greedy_mls)
(in module kernel_tuner.strategies.pso)
(in module kernel_tuner.strategies.random_sample)
(in module kernel_tuner.strategies.simulated_annealing)
tune_kernel() (in module kernel_tuner)
two_point_crossover() (in module kernel_tuner.strategies.genetic_algorithm)
U
uniform_crossover() (in module kernel_tuner.strategies.genetic_algorithm)
unscale_and_snap_to_nearest() (in module kernel_tuner.strategies.common)
W
weighted_choice() (in module kernel_tuner.strategies.genetic_algorithm)
write_file() (in module kernel_tuner.util)