Christian Mayr

TU Dresden is advertising a Research Associate / PhD Student position at the Chair of Highly-Parallel VLSI Systems and Neuro-Microelectronics within the Faculty of Electrical and Computer Engineering. The project is part of Horizon Europe OptimAIse and focuses on efficient language models , GenAI , and AI hardware deployment on SpiNNaker2 , a massively parallel European hardware platform developed at TU Dresden and commercialized by SpiNNcloud. The research aims to optimize large language models for inference and deployment using an existing software stack and ML compilers such as MLIR , with additional exploration of approaches like Mixture-of-Experts and other sparse, communication-avoiding model designs. The successful candidate will conduct scientific research on efficient language models and hardware-aware deployment, develop and train sparse GenAI models for SpiNNaker2, implement model layers on hardware, and publish results in top-tier conferences and journals. The work also contributes to requirements and recommendations for next-generation AI hardware such as SpiNNaker3. Eligibility: applicants should hold a Master’s degree or equivalent in computer science, electrical engineering, machine learning, or a related field. Strong programming skills in C++ and Python, good English, and the ability to work independently and in teams are required. Experience with LLVM/MLIR, embedded software, accelerator architectures, and parallel/distributed computing is an advantage. Funding: the position is funded (E 13 TV-L) and offers the chance to obtain a PhD. The role starts as soon as possible and is limited until May 31, 2029. Application deadline: October 20, 2026. Apply with a cover letter, CV, and degree certificates, preferably via the TUD SecureMail Portal, quoting reference code HPSN_OptimAIse_2026.