I recently explored CUDA programming on an embedded Jetson platform and developed a program to retrieve and display detailed GPU device properties. This experiment provided valuable insights into CUDA driver/runtime versions, memory architecture, multiprocessor configurations, and threading capabilities of the NVIDIA Tegra X1.
This learning journey was greatly enhanced by attending the Heterogeneous Parallel Programming (HPP) using CUDA and OpenCL seminar at AstroMediComp. A special thanks to Dr. Vijay D. Gokhale sir for his insightful lectures and guidance, which deepened my understanding of GPGPU computing.
💡 Key Takeaways from My Experiment:
✔ CUDA Driver & Runtime Version: 10.2
✔ GPU Compute Capability: 5.3
✔ Total Global Memory: 3.87 GB
✔ Number of SMs: 1
✔ Max Threads per SM: 2048
✔ Max Grid & Thread Dimensions explored
Excited to continue exploring CUDA for embedded systems and real-time GPU acceleration! 🔥
#CUDA #EmbeddedSystems #NVIDIAJetson #ParallelComputing #GPGPU #AstroMediComp #HPC #DeepLearning #EdgeAI