Release v13.3 of the CUDA samples with CUDA 13.3 Toolkit (#435)

This is the release of the CUDA 13.3 samples, which include additions for CUDA Tile C++, and updated CCCL and Python samples.

python/2_CoreConcepts/cudaGraphs/README.md

@@ -62,8 +62,8 @@ rebuilding.
 - CUDA Toolkit 13.0 or newer (matches `cuda-python` 13.x)
 - Python 3.10 or newer
 - `cuda-python` (>=13.0.0)
-- `cuda-core` (>=0.6.0)
-- `cupy-cuda13x` (>=13.0.0)
+- `cuda-core` (>=1.0.0)
+- `cupy-cuda13x` (>=14.0.0)
 
 ## Installation
 
@@ -77,8 +77,8 @@ pip install -r requirements.txt
 The `requirements.txt` installs:
 
 - `cuda-python` (>=13.0.0)
-- `cuda-core` (>=0.6.0)
-- `cupy-cuda13x` (>=13.0.0)
+- `cuda-core` (>=1.0.0)
+- `cupy-cuda13x` (>=14.0.0)
 
 ## How to Run
 

python/2_CoreConcepts/cudaGraphs/cudaGraphs.py

@@ -182,7 +182,7 @@ def main() -> int:
     stream = device.create_stream()
     # Tell CuPy to order its allocations on our stream so buffer initialization
     # below is serialized with the kernels we launch.
-    cp.cuda.ExternalStream(int(stream.handle)).use()
+    cp.cuda.Stream.from_external(stream).use()
 
     graph_builder = graph = None
     try:
@@ -214,9 +214,9 @@ def main() -> int:
         t_individual = run_pipeline_individual(
             stream, kernels, config, buffers, N, n_iters=args.iters
         )
-        assert cp.allclose(r3, expected, rtol=1e-5, atol=1e-5), (
-            "Individual pipeline produced incorrect results"
-        )
+        assert cp.allclose(
+            r3, expected, rtol=1e-5, atol=1e-5
+        ), "Individual pipeline produced incorrect results"
         print(
             f"\nIndividual launches: {args.iters} iters in {t_individual:.4f}s"
             f"  ({t_individual * 1e6 / args.iters:.2f} us/iter)"
@@ -228,9 +228,9 @@ def main() -> int:
 
         run_pipeline_graph(stream, graph, n_iters=5)  # warm up
         t_graph = run_pipeline_graph(stream, graph, n_iters=args.iters)
-        assert cp.allclose(r3, expected, rtol=1e-5, atol=1e-5), (
-            "Graph pipeline produced incorrect results"
-        )
+        assert cp.allclose(
+            r3, expected, rtol=1e-5, atol=1e-5
+        ), "Graph pipeline produced incorrect results"
         print(
             f"Graph replay:       {args.iters} iters in {t_graph:.4f}s"
             f"  ({t_graph * 1e6 / args.iters:.2f} us/iter)"

python/2_CoreConcepts/cudaGraphs/requirements.txt

@@ -1,5 +1,5 @@
 # CUDA Graphs Sample Requirements
 
 cuda-python>=13.0.0
-cuda-core>=0.6.0
-cupy-cuda13x>=13.0.0
+cuda-core>=1.0.0
+cupy-cuda13x>=14.0.0