PhD Position: Digital AI Chip Design for High Speed RF Systems

Join the ERC Synergy DISRUPT project! Build an energy-efficient digital AI/ML accelerator for real-time transmitter error correction in high-speed RF systems. Co-design and validate on real silicon.

Job description

Wireless networks are becoming increasingly power-hungry, and a significant portion of that energy is consumed by the transmitter. The ERC Synergy project DISRUPT is taking a bold step: moving RF-power generation further into the digital domain and enabling new transmitter concepts, but only if we can detect imperfections and correct them in real-time.

In this PhD, you will build the digital “brain” that makes this possible: an energy-efficient, silicon-proven AI/ML accelerator for transmitter error correction (digital predistortion/calibration). Your work will sit at the intersection of machine learning, DSP, and digital IC design, and you will validate your ideas on real chips alongside an international team; thus, it is also critical for you to be a team player, communicating smoothly with your colleagues.

Your main activities will include:

• Translate ML-based error-correction / DPD algorithms into hardware-friendly forms (model reduction, sparsity, quantization, fixed-point design).
• Design the architecture and RTL of a low-power accelerator that meets tight throughput/latency constraints.
• Take design through synthesis, layout, tape-out, and support PCB bring-up and measurements.
• Characterize the silicon in the lab: error-correction performance, throughput, latency, and power.
• Contribute to joint demonstrators and publications with project partners.

You will be mentored day-to-day by Dr. Chang Gao (TU Delft) and co-supervised by Prof. Anding Zhu (University College Dublin) and Prof. Leo de Vreede (TU Delft). This position directly connects to DISRUPT’s planned work on efficient ML-based DPD, accelerator architecture, silicon tape-out, and prototype measurements.

Electronic Circuits and Architectures (ELCA)

You will be part of the Electronic Circuits and Architectures (ELCA) section within the Department of Microelectronics at TU Delft. ELCA leads research in integrated circuit design, enhancing power efficiency, frequency performance, and functionality across RF, mm-wave, and wireless communication systems. The group bridges fundamental science with practical applications, pushing the boundaries of analog and digital circuit architectures.

You will have access to state-of-the-art design and testing facilities, including advanced fabrication technologies and specialized measurement labs. As part of a dynamic and interdisciplinary team, you will collaborate with experts in circuit design, signal processing, and emerging semiconductor technologies.

Job requirements

Essential Requirements:

• MSc degree in Electrical / Electronic / Computer Engineering, or a closely related field.
• Solid background in digital IC design and digital signal processing.
• Hands-on RTL design skills (SystemVerilog / Verilog / VHDL) plus scripting (Python / MATLAB / C/C++).
• Strong command of English.
• Strong team player with excellent communication skills to collaborate smoothly with international colleagues.

Nice to have:

• Practical experience with machine-learning frameworks (e.g., PyTorch).
• Prior tape-out experience (ASIC or a complex FPGA prototype) and familiarity with the digital back-end flow (synthesis, P&R, timing closure).
• Knowledge of low-power digital design techniques (e.g., clock/power gating, DVFS).
• A strong curiosity about RF transmitters, DPD algorithms, and analog/mixed-signal challenges.

Conditions of employment

Doctoral candidates will be offered a 4-year period of employment in principle, but in the form of 2 employment contracts. An initial 1.5 year contract with an official go/no go progress assessment within 15 months. Followed by an additional contract for the remaining 2.5 years assuming everything goes well and performance requirements are met.