NVIDIA/nvbench
1
0
Fork 0
mirror of https://github.com/NVIDIA/nvbench.git synced 2026-05-12 17:25:41 +00:00
Code Issues Packages Projects Releases Wiki Activity
748 Commits 17 Branches 5 Tags
93bc59d05c7d073a16e6f24d01ebfb4fbd92c744
Commit Graph

119 Commits

Author SHA1 Message Date
Oleksandr Pavlyk
93bc59d05c Renamed CUTLASS example to reflect that it uses CuteDSL 2026-04-01 08:24:29 -05:00
Oleksandr Pavlyk
e4cfddeb87 Rewrote cutlass_gemm example to use CuteDSL 2026-04-01 08:23:41 -05:00
Oleksandr Pavlyk
3f284b4004 Renamed cccl_* examples 
cccl_parallel_* -> cuda_compute_*
cccl_cooperative_* -> cuda_coop_*
 2026-04-01 08:20:20 -05:00
Oleksandr Pavlyk
5bdb30f4b6 Update to cccl_parallel_segmented_reduce example per changes in API 
Update namespace changes. Use make_segmented_reduce factory function,
and update call signatures.
 2026-04-01 08:18:15 -05:00
Oleksandr Pavlyk
d8739fc208 Update to cccl_cooperative_block_reduce example 2026-04-01 08:17:52 -05:00
Oleksandr Pavlyk
974eb5ee0f Replace use of cupy.cuda.ExternalStream with cupy.cuda.Stream.from_external 2026-04-01 08:17:12 -05:00
Oleksandr Pavlyk
7c60edcc0a cuda.core.experimental -> cuda.core 2026-04-01 08:16:04 -05:00
Oleksandr Pavlyk
836a6c12f4 Merge pull request #326 from oleksandr-pavlyk/fix-sfinae-incomplete 
Fix GCC16 sfinae incomplete warnings.

GCC16 started requiring that the type `T` used in `std::reference_wrapper<T>` is complete where using `-std=c++17`. Since NVBench has to forward declare some types in header files to break circular dependency, use of incomplete type breaks build due to use of `-Werror` flag due to `-Wsfinae-incomplete` warning emitted by GCC16.

This commit replaced affected uses of `std::reference_wrapper<const nvbench::benchmark_base>` in state.cxx, and `std::reference_wrapper<nvbench::printer_base>` in benchmark_base.cxx with raw pointers.
 2026-03-24 16:02:28 -05:00
Bernhard Manfred Gruber
4164909c52 Feedback 2026-02-28 01:19:18 +01:00
Bernhard Manfred Gruber
0abc8ec82b Extend nvbench_compare.py with --plot, axis/benchmark filtering, and dark mode 
Co-authored-by: Oleksandr Pavlyk <21087696+oleksandr-pavlyk@users.noreply.github.com>
 2026-02-27 11:06:20 +01:00
Bernhard Manfred Gruber
800f640c20 Apply reviewer feedback 2026-02-26 19:23:51 +01:00
Bernhard Manfred Gruber
d3a0bec4a8 Feedback from review 2026-02-05 14:13:16 +01:00
Bernhard Manfred Gruber
28ed32bb47 Implement dark mode using style sheets 2026-02-05 14:00:33 +01:00
Bernhard Manfred Gruber
ec9759037d I have no idea what I am doing 2026-02-05 11:15:27 +01:00
Bernhard Manfred Gruber
ccde9fc4d4 More 2026-02-05 10:56:36 +01:00
Bernhard Manfred Gruber
0be190b407 Add a script to plot benchmark results 2026-02-05 10:36:52 +01:00
Nader Al Awar
dc59f98ecd Remove cupti from cuda-bench dependencies (#311) 2026-02-03 14:16:26 -06:00
Bernhard Manfred Gruber
c6ef87575c Allow partial comparison in nvbench_compare.py 
Fixes: #295
 2026-02-03 16:32:11 +01:00
Nader Al Awar
d75fc74162 Merge branch 'main' into remove-cupti-python 2026-02-03 08:58:41 -06:00
Nader Al Awar
4fa4296810 Remove cuda.pathfinder function 2026-02-02 16:43:45 -06:00
Nader Al Awar
f2d5730104 Disable CUPTI in cmake file 2026-02-02 16:03:15 -06:00
Nader Al Awar
6df5fc8c67 Remove cupti from cuda-bench dependencies 2026-02-02 15:37:13 -06:00
Oleksandr Pavlyk
8ff0557ad8 Replace use of py::handle to store global_registry 
Use py::gil_safe_call_once_and_store facility pybind11 provides.
 2026-02-02 11:55:48 -06:00
Oleksandr Pavlyk
39c29026fd Move docstrings from PYI file to implementation 
Added tests that docstrings exist and are not empty.

This closes #291
 2026-02-02 11:55:48 -06:00
Nader Al Awar
edf0b80599 Add installation instructions 2026-01-30 09:32:44 -06:00
Nader Al Awar
fa1eed69c0 Rename test file to refer to cuda_bench 2026-01-29 13:53:29 -06:00
Nader Al Awar
711c1e2eb1 Replace all occurences of pynvbench with cuda-bench 2026-01-29 13:25:17 -06:00
Nader Al Awar
5e7adc5c3f Build multi architecture cuda wheels (#302) 
* Add cuda architectures to build wheel for

* Package scripts in wheel

* Separate cuda major version extraction to fix architecutre selection logic

* Add back statement printing cuda version

* [pre-commit.ci] auto code formatting

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
 2026-01-29 01:13:24 +00:00
Ashwin Srinath
a681e2185d Add multi-cuda wheel build (#289) 
Co-authored-by: Ashwin Srinath <shwina@users.noreply.github.com>
Co-authored-by: Nader Al Awar <naderalawar@gmail.com>
 2026-01-28 10:37:55 -05:00
Oleksandr Pavlyk
f6a9b245d3 Only trigger skipping of outstanding benchmarks on KeyboardInterrupt exception, on others benchmakr is to continue execution 2025-12-08 14:46:59 -06:00
Oleksandr Pavlyk
7e9a9a8983 Replace main_arg_run_benchmarks with run_interriptible 
This loop uses benchmark.run_or_skip to resolve #284 even
for scripts that contain more than one benchmark, or when
a script with a single benchmark is executed when more than
one device is available.
 2025-12-08 14:29:27 -06:00
Oleksandr Pavlyk
a7763bdd7a Remove debug outputs 2025-12-08 12:25:31 -06:00
Oleksandr Pavlyk
b2a80c92b8 Revert "Scripts to triage 284" 
This reverts commit c286199adc.
 2025-12-08 11:53:08 -06:00
Oleksandr Pavlyk
ce9a76167f Use nvbench::stop_runner_loop to signal stop of runner loop 
Add try/catch around Python calls to improve keyboard interrup
response.
 2025-12-05 19:38:11 -06:00
Oleksandr Pavlyk
c286199adc Scripts to triage 284 2025-12-05 19:38:11 -06:00
Oleksandr Pavlyk
de471e1d42 Use pybind11==3.0.1, do not use pybind11_add_module 2025-12-05 19:38:11 -06:00
Ashwin Srinath
77b7afc3c9 Remove the Python version file 2025-12-03 16:23:14 -05:00
Ashwin Srinath
29389b5791 Initial wheel build and publishing infrastructure 2025-12-03 10:15:32 -05:00
Asher Mancinelli
e91559edf0 Update README.md 2025-11-14 14:34:18 -08:00
Jaya Venkatesh
0f997271f7 added numba-cuda to requirements 
Signed-off-by: Jaya Venkatesh <jjayabaskar@nvidia.com>
 2025-09-16 14:54:08 -07:00
Jaya Venkatesh
bfa6a6c7c6 remove pynvjitlink references in examples 
Signed-off-by: Jaya Venkatesh <jjayabaskar@nvidia.com>
 2025-09-08 16:00:19 -07:00
Oleksandr Pavlyk
b5e4b4ba31 cuda.nvbench -> cuda.bench 
Per PR review suggestion:
   - `cuda.parallel`    - device-wide algorithms/Thrust
   - `cuda.cooperative` - Cooperative algorithsm/CUB
   - `cuda.bench`       - Benchmarking/NVBench
 2025-08-04 13:42:43 -05:00
Oleksandr Pavlyk
c2a2acc9b6 Change float64_t arg-type for set_throttle_threshold to float32_t 
The C++ method signature of set_throttle_threshold/set_trottle_recovery_delay,
which uses nvbench::float32_t
 2025-08-04 12:14:52 -05:00
Oleksandr Pavlyk
584f48ac97 Remove warm-up invocations outside of launcher in examples/throughout and auto_throughput 2025-08-04 12:14:44 -05:00
Oleksandr Pavlyk
d8b0acc8d4 Export exception to nvbench namespace 2025-08-04 12:00:42 -05:00
Oleksandr Pavlyk
9dfdd8af89 Minimal test file 2025-08-04 11:59:17 -05:00
Oleksandr Pavlyk
6aff4712f8 Change permissions of test/run_1.py 2025-08-04 10:13:08 -05:00
Oleksandr Pavlyk
73e18419b2 Stub of __cuda_stream__ special method declare tuple[int, int] as return type 
This is to indicate that special method always returns a pair of integers
 2025-08-04 10:11:33 -05:00
Oleksandr Pavlyk
a5e0a48f80 Add test test functions for cpp/python exceptions 2025-08-04 10:09:10 -05:00
Oleksandr Pavlyk
40a2337a6b Review fix: make nvbenhch_run_error constructable 
Allow `throw nvbench_run_error("Msg");` to compile.

Add comment around definition of nvbench_run_error
 2025-08-04 10:09:04 -05:00