I’m a CUDA/GPU trainer and DSP/wireless engineer who helps developers move from CPU to GPU with practical, profiler-driven workflows. Over the past 14+ years I’ve built and optimized signal-processing and wireless PHY algorithms (LTE, NB-IoT, 5G/NTN) and accelerated AI/HPC workloads using C/C++/CUDA, cuBLAS/cuFFT, and Nsight tools.

How I teach

Lab-first, profile-first: write → measure → analyze → optimize

Zero-setup learning with Google Colab (no local GPU required)

Short, focused videos, runnable notebooks, and realistic mini-projects

What you’ll build

Your first CUDA kernels with correct indexing & grid-striding

CPU vs GPU benchmarks with real speedup targets

Foundations you can extend to AI/LLM, image/signal processing, or HPC