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.
This commit is contained in:
Dheemanth
2026-05-27 14:50:59 -07:00
committed by GitHub
parent aeab82ff30
commit b7c5481c55
150 changed files with 7977 additions and 246 deletions

View File

@@ -28,7 +28,7 @@ This sample demonstrates CuPy integration with cuda.core streams:
stream = device.create_stream()
# Use with CuPy operations
cp.cuda.ExternalStream(int(stream.handle)).use()
cp.cuda.Stream.from_external(stream).use()
```
## Key APIs
@@ -43,7 +43,7 @@ cp.cuda.ExternalStream(int(stream.handle)).use()
- `cp.fft.rfft()` - Real-to-complex FFT (GPU-accelerated via cuFFT)
- `cp.fft.rfftfreq()` - Generate frequency bins for rfft
- `cp.cuda.ExternalStream()` - Interop with cuda.core streams
- `cp.cuda.Stream.from_external()` - Interop with cuda.core streams
### From NumPy:
@@ -115,7 +115,7 @@ VERIFICATION
GPU vs CPU FFT magnitude: Test PASSED
Frequency Detection Accuracy:
440 Hz: ✓
440 Hz: [OK]
...
Done

View File

@@ -180,7 +180,7 @@ def run_fft_analysis(
print(f"Compute Capability: sm_{device.arch}")
# Make CuPy use our cuda.core stream
cp.cuda.ExternalStream(int(stream.handle)).use()
cp.cuda.Stream.from_external(stream).use()
# Define test signal: composite of multiple frequencies
test_frequencies = [440.0, 880.0, 1320.0, 2000.0, 5000.0] # Hz
@@ -208,7 +208,7 @@ def run_fft_analysis(
print("GPU FFT (cuFFT)")
print("-" * 60)
event_opts = EventOptions(enable_timing=True)
event_opts = EventOptions(timing_enabled=True)
# Warmup
d_fft_result = cp.fft.rfft(d_signal)
@@ -291,7 +291,7 @@ def run_fft_analysis(
all_found = True
for expected_freq in test_frequencies:
found = any(abs(f - expected_freq) < 10 for f in detected_freqs)
status = "✓" if found else "✗"
status = "[OK]" if found else "[FAIL]"
print(f" {expected_freq:6.0f} Hz: {status}")
all_found = all_found and found

View File

@@ -1,6 +1,6 @@
# FFT Signal Analysis 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
numpy>=2.3.2