What I’m doing now
GPU Computing
I first got interested in GPUs when I was building my first PC, I remember seeing the name NVIDIA for the first time back then and buying an RTX 2060 with the only goal being running games at high quality.
Now, I am researching anything GPU, from GPU microarchitecture to programming on GPUs using CUDA. I am following several resources but I reccommend these to get started:
- PMPP (Programming Massively Parallel Processors) - book/video lectures
- NVIDIA whitepapers - great to break down the microarchitecture and design decisions for different chips
- CUDA programming guide - official NVIDIA docs
- CUDA by Example - book
AI workloads
I first explored AI, specifically computer vision applications, during my university degree. But it then got pushed into the backburner for a long while, until very recently, when I started exploring how GPU computing is transforming the way we approach and solve problems.
This renewed focus on AI is like welcoming an old friend. Right now I am researching 2 specific domains in AI:
- GNNs as I find the graph structure fascinating as it can represent so many aspects of the world around us.
- Computer Vision, where you a computer learns how to perceive the world around it.
Computer Architecture
You can’t really understand GPUs without understanding the underlying architecture behind them. So I am going to explore the foundations of computing from the silicon upwards. I am using the following resources to guide me:
- Digital Design and Computer Architecture - course by Onur Mutlu available on
youtube
- This also includes Hardware Description Languages to actually implement the things I learn about. (right now I just have ICarus Verilog (iverilog) working…)
- Computer Architecture: A Quantitative Approach - book
You can follow my progress and detailed notes on these topics on the notes page
Last modified: 30/08/2025