Personal blog

Through my Master Thesis, SANDA, I explored two interesting topics in the field of Chip Design and AI accelerators: Block-Floating Point (BFP) and Bit-Serial Architectures. These two concepts, while seemingly distinct, share a common goal of optimizing computational efficiency and resource utilization in hardware design. [Read More]

As I wrap up this academic cycle, I wanted to reflect on three distinct projects that I conducted at KU Leuven and how they managed to make me a better, smarter and more versatile future engineer. I conducted many projects this last year, Let’s dive in! [Read More]

As Large Language Models (LLMs) continue to grow in size and capability, deploying them efficiently has become increasingly challenging. Enter Auto-ADPQ (Adaptive Post-Training Quantization) - a Python package I developed that implements a zero-shot, calibration-free method for quantizing LLMs to 4-bit precision while maintaining model quality. [Read More]

In this blogpost, we explore the multiple design steps we went through to design a Co-Design implementation of the RSA algorithm using the Pynq Z-2 and C code combined with ARM assembly. We will motivate our design choices and provide some insights and metrics about the implementation. Finally, we will... [Read More]

After finishing my internship at NXP, I took a break… until my brain couldn’t handle the boredom. So I pulled back my long list of ideas to make me a billionaire (spoiler: the ideas are all trash but fun). Among the ideas was: “Make a simple conversion tool using Pandoc”.... [Read More]