Computer Engineer | FPGA & Digital Design | Embedded Systems
LinkedIn: https://linkedin.com/in/andreea-anastasiu Email: aanastasiu52@gmail.com
I'm a Computer Engineer (BSc, Technical University of Cluj-Napoca) currently pursuing an MSc in Modern Signal Processing Techniques at TU Iași. I work close to the hardware — building efficient solutions in Verilog/VHDL and exploring the interaction between software and programmable logic.
In my Bachelor's thesis, I designed and compared five CNN inference paradigms on a Zynq-7000 SoC (Zybo Z7-20): a custom VHDL hardware accelerator, a scalar C baseline, ARM NEON SIMD, and a FINN-generated dataflow accelerator — reaching inference latency as low as 2.75 ms. This involved AXI DMA, Vivado block design, and bare-metal Vitis development.
I completed a Summer Internship at Analog Devices, working across Hardware, HDL, Embedded Linux, DevOps and Software modules — including a custom PCB design in KiCad, an FPGA module in Verilog, an ADC Linux driver, and a Python GUI for real-time motion visualization.
Currently building on FPGA/digital design foundations through my Master's studies. Interested in digital design, verification, hardware acceleration, signal processing, and embedded Linux. Open to internship and entry-level opportunities in FPGA/embedded systems.
CNN Inference Acceleration on FPGA (Bachelor Thesis) Designed, implemented, and benchmarked 4 CNN inference approaches on a Zynq-7000 SoC (VHDL hardware accelerator, FINN dataflow, scalar C, ARM NEON SIMD) using the same model/weights/dataset for a fair comparison.
| Implementation | Accuracy | Latency |
|---|---|---|
| FINN (dataflow) | 71.4% | 2.75 ms |
| VHDL accelerator | 82.2% | 31 ms |
| Scalar C | 82.2% | 249 ms |
| ARM NEON (SIMD) | 82.2% | 292 ms |
Stack: VHDL, AXI DMA, Vivado, Vitis, FINN/Brevitas, ARM NEON
Garden Robot – Autonomous Environmental Monitoring Mobile agricultural robot with CO2/temperature/humidity/soil-moisture sensing and autonomous ultrasonic obstacle avoidance (Arduino + Raspberry Pi 4). Stack: Arduino, Embedded Systems
- Aspects of Hamming Encryption Using Raspberry Pi Pico — AFASES 2025
- Arduino vs Raspberry Pi – Measuring and Interpreting Physical Quantities — Cygnus Journal, 2024
Verilog, VHDL, C, C++, Python, Xilinx Vivado/Vitis, Embedded Linux, Git