Skip to content

Use cuda::stream_ref in benchmarks and examples #23697

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
vyasr wants to merge 3 commits into from
Closed

Use cuda::stream_ref in benchmarks and examples #23697

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
96118dd
Use cuda::stream_ref in benchmarks and examples
vyasr Aug 17, 2026
6b0e3c8
Address stream_ref benchmark review comments
vyasr Aug 17, 2026
5bb58ce
Fix stream_ref benchmark formatting
vyasr Aug 18, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 1 addition & 3 deletions cpp/benchmarks/ast/transform.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand All @@ -17,8 +17,6 @@
#include <cudf/utilities/default_stream.hpp>
#include <cudf/utilities/error.hpp>

#include <rmm/cuda_stream_view.hpp>

#include <cuda/iterator>

#include <nvbench/nvbench.cuh>
Expand Down
6 changes: 3 additions & 3 deletions cpp/benchmarks/binaryop/binaryop.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand Down Expand Up @@ -104,7 +104,7 @@ static void BM_string_compare_binaryop_transform(nvbench::state& state)
auto const mem_stats_logger = cudf::memory_stats_logger();

state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
rmm::cuda_stream_view stream{launch.get_stream().get_stream()};
cuda::stream_ref stream{launch.get_stream().get_stream()};
std::unique_ptr<cudf::column> reduction =
cudf::binary_operation(table.get_column(0), table.get_column(1), cmp_op, bool_type, stream);
std::for_each(
Expand All @@ -115,7 +115,7 @@ static void BM_string_compare_binaryop_transform(nvbench::state& state)
table.get_column(idx * 2), table.get_column(idx * 2 + 1), cmp_op, bool_type, stream);
std::unique_ptr<cudf::column> reduced =
cudf::binary_operation(*comparison, *reduction, reduce_op, bool_type, stream);
stream.synchronize();
stream.sync();
reduction = std::move(reduced);
});
});
Expand Down
4 changes: 2 additions & 2 deletions cpp/benchmarks/binaryop/polynomials.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2025-2026, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2025-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand Down Expand Up @@ -58,7 +58,7 @@ static void BM_binaryop_polynomials(nvbench::state& state)
state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) {
// computes polynomials: (((ax + b)x + c)x + d)x + e... = ax**4 + bx**3 + cx**2 + dx + e....
cudf::benchmark::scoped_range range{"benchmark_iteration"};
rmm::cuda_stream_view stream{launch.get_stream().get_stream()};
cuda::stream_ref stream{launch.get_stream().get_stream()};
std::vector<std::unique_ptr<cudf::column>> intermediates;

auto result = cudf::make_column_from_scalar(constants[0], num_rows, stream);
Expand Down
6 changes: 3 additions & 3 deletions cpp/benchmarks/bitmask/bitmask_and.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand Down Expand Up @@ -70,7 +70,7 @@ void BM_segmented_bitmask_and(nvbench::state& state)

auto [segments, masks, mask_pointers, mask_begin_bits, data_bytes] = setup_masks(state);

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
state.add_element_count(data_bytes, "input size");
state.template add_global_memory_reads<nvbench::int8_t>(data_bytes);
auto const mem_stats_logger = cudf::memory_stats_logger();
Expand All @@ -91,7 +91,7 @@ void BM_multi_segment_bitmask_and(nvbench::state& state)

auto [segments, masks, mask_pointers, mask_begin_bits, data_bytes] = setup_masks(state);

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
state.add_element_count(data_bytes, "input size");
state.template add_global_memory_reads<nvbench::int8_t>(data_bytes);
auto const mem_stats_logger = cudf::memory_stats_logger();
Expand Down
10 changes: 5 additions & 5 deletions cpp/benchmarks/bitmask/set_null_mask.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand Down Expand Up @@ -61,7 +61,7 @@ void BM_setnullmask(nvbench::state& state)
rmm::device_buffer mask = cudf::create_null_mask(mask_size, cudf::mask_state::UNINITIALIZED);
auto begin = 0, end = mask_size;

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
auto const mem_stats_logger = cudf::memory_stats_logger();

state.exec(nvbench::exec_tag::sync | nvbench::exec_tag::timer,
Expand Down Expand Up @@ -90,7 +90,7 @@ void BM_setnullmask_unsafe_bulk(nvbench::state& state)
auto [begin_bits, end_bits, valids, masks, masks_ptr] =
generate_test_data(num_masks, mask_size, use_variable_mask_sizes);

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
auto const mem_stats_logger = cudf::memory_stats_logger();

state.exec(nvbench::exec_tag::sync | nvbench::exec_tag::timer,
Expand Down Expand Up @@ -118,7 +118,7 @@ void BM_setnullmask_safe_bulk(nvbench::state& state)
auto [begin_bits, end_bits, valids, masks, masks_ptr] =
generate_test_data(num_masks, mask_size, use_variable_mask_sizes);

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
auto const mem_stats_logger = cudf::memory_stats_logger();

state.exec(nvbench::exec_tag::sync | nvbench::exec_tag::timer,
Expand Down Expand Up @@ -146,7 +146,7 @@ void BM_setnullmask_loop(nvbench::state& state)
auto [begin_bits, end_bits, valids, masks, masks_ptr] =
generate_test_data(num_masks, mask_size, use_variable_mask_sizes);

state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().value()));
state.set_cuda_stream(nvbench::make_cuda_stream_view(cudf::get_default_stream().get()));
auto const mem_stats_logger = cudf::memory_stats_logger();

state.exec(nvbench::exec_tag::sync | nvbench::exec_tag::timer,
Expand Down
23 changes: 12 additions & 11 deletions cpp/benchmarks/common/ndsh_data_generator/ndsh_data_generator.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,9 +24,10 @@
#include <cudf/transform.hpp>
#include <cudf/unary.hpp>

#include <rmm/cuda_stream_view.hpp>
#include <rmm/resource_ref.hpp>

#include <cuda/stream>

#include <array>
#include <string>
#include <vector>
Expand Down Expand Up @@ -142,7 +143,7 @@ constexpr std::array vocab_containers{
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_orders_independent(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -266,7 +267,7 @@ std::unique_ptr<cudf::table> generate_orders_independent(double scale_factor,
*/
std::unique_ptr<cudf::table> generate_lineitem_partial(cudf::table_view const& orders_independent,
double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -437,7 +438,7 @@ std::unique_ptr<cudf::table> generate_lineitem_partial(cudf::table_view const& o
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_orders_dependent(cudf::table_view const& lineitem_partial,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -530,7 +531,7 @@ std::unique_ptr<cudf::table> generate_orders_dependent(cudf::table_view const& l
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_partsupp(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -578,7 +579,7 @@ std::unique_ptr<cudf::table> generate_partsupp(double scale_factor,
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_part(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -704,7 +705,7 @@ std::unique_ptr<cudf::table> generate_part(double scale_factor,
*/
std::tuple<std::unique_ptr<cudf::table>, std::unique_ptr<cudf::table>, std::unique_ptr<cudf::table>>
generate_orders_lineitem_part(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -771,7 +772,7 @@ generate_orders_lineitem_part(double scale_factor,
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_supplier(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -832,7 +833,7 @@ std::unique_ptr<cudf::table> generate_supplier(double scale_factor,
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_customer(double scale_factor,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -899,7 +900,7 @@ std::unique_ptr<cudf::table> generate_customer(double scale_factor,
* @param stream CUDA stream used for device memory operations and kernel launches
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_nation(rmm::cuda_stream_view stream,
std::unique_ptr<cudf::table> generate_nation(cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -939,7 +940,7 @@ std::unique_ptr<cudf::table> generate_nation(rmm::cuda_stream_view stream,
* @param stream CUDA stream used for device memory operations and kernel launches
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_region(rmm::cuda_stream_view stream,
std::unique_ptr<cudf::table> generate_region(cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down
14 changes: 7 additions & 7 deletions cpp/benchmarks/common/ndsh_data_generator/ndsh_data_generator.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
/*
* SPDX-FileCopyrightText: Copyright (c) 2024-2025, NVIDIA CORPORATION.
* SPDX-FileCopyrightText: Copyright (c) 2024-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
* SPDX-License-Identifier: Apache-2.0
*/

Expand All @@ -22,7 +22,7 @@ namespace datagen {
std::tuple<std::unique_ptr<cudf::table>, std::unique_ptr<cudf::table>, std::unique_ptr<cudf::table>>
generate_orders_lineitem_part(
double scale_factor,
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

/**
Expand All @@ -34,7 +34,7 @@ generate_orders_lineitem_part(
*/
std::unique_ptr<cudf::table> generate_partsupp(
double scale_factor,
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

/**
Expand All @@ -46,7 +46,7 @@ std::unique_ptr<cudf::table> generate_partsupp(
*/
std::unique_ptr<cudf::table> generate_supplier(
double scale_factor,
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

/**
Expand All @@ -58,7 +58,7 @@ std::unique_ptr<cudf::table> generate_supplier(
*/
std::unique_ptr<cudf::table> generate_customer(
double scale_factor,
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

/**
Expand All @@ -68,7 +68,7 @@ std::unique_ptr<cudf::table> generate_customer(
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_nation(
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

/**
Expand All @@ -78,7 +78,7 @@ std::unique_ptr<cudf::table> generate_nation(
* @param mr Device memory resource used to allocate the returned column's device memory
*/
std::unique_ptr<cudf::table> generate_region(
rmm::cuda_stream_view stream = cudf::get_default_stream(),
cuda::stream_ref stream = cudf::get_default_stream(),
rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());

} // namespace datagen
Expand Down
24 changes: 12 additions & 12 deletions cpp/benchmarks/common/ndsh_data_generator/random_column_generator.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,7 +79,7 @@ struct random_number_generator {
std::unique_ptr<cudf::column> generate_random_string_column(cudf::size_type lower,
cudf::size_type upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand Down Expand Up @@ -107,7 +107,7 @@ template <typename T>
std::unique_ptr<cudf::column> generate_random_numeric_column(T lower,
T upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand All @@ -127,33 +127,33 @@ template std::unique_ptr<cudf::column> generate_random_numeric_column<int8_t>(
int8_t lower,
int8_t upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

template std::unique_ptr<cudf::column> generate_random_numeric_column<int16_t>(
int16_t lower,
int16_t upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

template std::unique_ptr<cudf::column> generate_random_numeric_column<cudf::size_type>(
cudf::size_type lower,
cudf::size_type upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

template std::unique_ptr<cudf::column> generate_random_numeric_column<double>(
double lower,
double upper,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

std::unique_ptr<cudf::column> generate_primary_key_column(cudf::scalar const& start,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand All @@ -162,7 +162,7 @@ std::unique_ptr<cudf::column> generate_primary_key_column(cudf::scalar const& st

std::unique_ptr<cudf::column> generate_repeat_string_column(std::string const& value,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand All @@ -173,7 +173,7 @@ std::unique_ptr<cudf::column> generate_repeat_string_column(std::string const& v
std::unique_ptr<cudf::column> generate_random_string_column_from_set(
cudf::host_span<char const* const> set,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand All @@ -199,7 +199,7 @@ template <typename T>
std::unique_ptr<cudf::column> generate_repeat_sequence_column(T seq_length,
bool zero_indexed,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr)
{
CUDF_BENCHMARK_RANGE();
Expand All @@ -224,14 +224,14 @@ template std::unique_ptr<cudf::column> generate_repeat_sequence_column<int8_t>(
int8_t seq_length,
bool zero_indexed,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

template std::unique_ptr<cudf::column> generate_repeat_sequence_column<cudf::size_type>(
cudf::size_type seq_length,
bool zero_indexed,
cudf::size_type num_rows,
rmm::cuda_stream_view stream,
cuda::stream_ref stream,
rmm::device_async_resource_ref mr);

} // namespace cudf::datagen
Loading
Loading