NVIDIA · mdboom · Mar 20, 2026 · Mar 20, 2026 · Mar 20, 2026 · Mar 20, 2026
diff --git a/.github/workflows/test-wheel-linux.yml b/.github/workflows/test-wheel-linux.yml
@@ -96,12 +96,15 @@ jobs:
         uses: nv-gha-runners/setup-proxy-cache@main
         continue-on-error: true
 
+      - name: Install uv
+        uses: astral-sh/setup-uv@v7
+
       - name: Install dependencies
         uses: ./.github/actions/install_unix_deps
         continue-on-error: false
         with:
-          # for artifact fetching, graphics libs
-          dependencies: "jq wget libgl1 libegl1"
+          # for artifact fetching, graphics libs, g++ required for cffi in example
+          dependencies: "jq wget libgl1 libegl1 g++"
           dependent_exes: "jq wget"
 
       - name: Set environment variables

diff --git a/cuda_core/examples/cuda_graphs.py b/cuda_core/examples/cuda_graphs.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -10,6 +10,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
 import sys
 import time
 
@@ -121,6 +125,9 @@ def main():
         end_time = time.time()
 
         graph_execution_time = end_time - start_time
+        if graph_execution_time == 0.0:
+            print("Graph execution time is too fast to measure accurately.")
+            graph_execution_time = 1e-9  # Assign a small value to avoid division by zero in speedup calculation
         print(f"Graph execution time: {graph_execution_time:.6f} seconds")
 
         # Verify results

diff --git a/cuda_core/examples/gl_interop_plasma.py b/cuda_core/examples/gl_interop_plasma.py
@@ -53,9 +53,10 @@
 # effect popular in the demoscene). The window title shows the current FPS.
 # Close the window or press Escape to exit.
 #
-# Requirements
-# ============
-#   pip install pyglet
+
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core>0.6.0", "pyglet"]
+# ///
 
 import ctypes
 import sys

diff --git a/cuda_core/examples/jit_lto_fractal.py b/cuda_core/examples/jit_lto_fractal.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -12,6 +12,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
 import argparse
 import sys
 

diff --git a/cuda_core/examples/memory_ops.py b/cuda_core/examples/memory_ops.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -10,6 +10,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
 import sys
 
 import cupy as cp

diff --git a/cuda_core/examples/pytorch_example.py b/cuda_core/examples/pytorch_example.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "torch"]
+# ///
+
 import sys
 
 import torch

diff --git a/cuda_core/examples/saxpy.py b/cuda_core/examples/saxpy.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -10,8 +10,17 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
+
 import sys
 
+from cuda import pathfinder
+
+print(pathfinder.load_nvidia_dynamic_lib("nvrtc"))
+
 import cupy as cp
 
 from cuda.core import Device, LaunchConfig, Program, ProgramOptions, launch

diff --git a/cuda_core/examples/show_device_properties.py b/cuda_core/examples/show_device_properties.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2025-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core"]
+# ///
+
 import sys
 
 from cuda.core import Device, system

diff --git a/cuda_core/examples/simple_multi_gpu_example.py b/cuda_core/examples/simple_multi_gpu_example.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "cupy-cuda13x"]
+# ///
+
 import sys
 
 import cupy as cp

diff --git a/cuda_core/examples/strided_memory_view_cpu.py b/cuda_core/examples/strided_memory_view_cpu.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "cffi", "setuptools"]
+# ///
+
 import importlib
 import string
 import sys

diff --git a/cuda_core/examples/strided_memory_view_gpu.py b/cuda_core/examples/strided_memory_view_gpu.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
 import string
 import sys
 

diff --git a/cuda_core/examples/thread_block_cluster.py b/cuda_core/examples/thread_block_cluster.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -10,6 +10,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core"]
+# ///
+
 import os
 import sys
 
@@ -67,6 +71,7 @@ def main():
 
     cuda_path = os.environ.get("CUDA_PATH", os.environ.get("CUDA_HOME"))
     if cuda_path is None:
+        print(os.environ)
         print("this example requires a valid CUDA_PATH environment variable set", file=sys.stderr)
         sys.exit(1)
     cuda_include = os.path.join(cuda_path, "include")

diff --git a/cuda_core/examples/tma_tensor_map.py b/cuda_core/examples/tma_tensor_map.py
@@ -22,6 +22,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core>0.6.0", "cupy-cuda13x"]
+# ///
+
 import os
 import sys
 

diff --git a/cuda_core/examples/vector_add.py b/cuda_core/examples/vector_add.py
@@ -1,4 +1,4 @@
-# SPDX-FileCopyrightText: Copyright (c) 2024-2025 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2024-2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # SPDX-License-Identifier: Apache-2.0
 
@@ -9,6 +9,10 @@
 #
 # ################################################################################
 
+# /// script
+# dependencies = ["cuda_bindings", "cuda_core", "nvidia-cuda-nvrtc", "cupy-cuda13x"]
+# ///
+
 import cupy as cp
 
 from cuda.core import Device, LaunchConfig, Program, ProgramOptions, launch

diff --git a/cuda_core/pyproject.toml b/cuda_core/pyproject.toml
@@ -56,7 +56,7 @@ cu12 = ["cuda-bindings[all]==12.*"]
 cu13 = ["cuda-bindings[all]==13.*"]
 
 [dependency-groups]
-test = ["cython>=3.2,<3.3", "setuptools", "pytest>=6.2.4", "pytest-randomly", "pytest-repeat", "pytest-rerunfailures"]
+test = ["cython>=3.2,<3.3", "setuptools", "pytest>=6.2.4", "pytest-randomly", "pytest-repeat", "pytest-rerunfailures", "cffi"]
 ml-dtypes = ["ml-dtypes>=0.5.4,<0.6.0"]
 test-cu12 = [ {include-group = "ml-dtypes" }, {include-group = "test" }, "cupy-cuda12x; python_version < '3.14'", "cuda-toolkit[cudart]==12.*"]  # runtime headers needed by CuPy
 test-cu13 = [ {include-group = "ml-dtypes" }, {include-group = "test" }, "cupy-cuda13x; python_version < '3.14'", "cuda-toolkit[cudart]==13.*"]  # runtime headers needed by CuPy