Krithik Sama

I like working on problems where software abstractions eventually have to confront real hardware constraints. A model may look great in Python, but fitting it onto an FPGA, meeting timing, reducing resource use, or making it run efficiently on a CPU is where the engineering gets interesting to me.

That interest has pulled me into projects across FPGA acceleration, RTL design and verification, model quantization, design-space exploration, and optimized AI inference. I've worked on everything from mapping neural networks directly into FPGA fabric to exploring hardware-friendly model architectures.

What I enjoy most is the iteration: build something, measure it, find the bottleneck, and rethink the design around what the hardware is actually telling you. That's the kind of work I want to keep doing — especially in AI hardware, digital design, and hardware/software co-design.