Overview / Requirements
Master's thesis topics are available in several key areas of modern artificial intelligence and AI hardware, including: 1. AI Security 2. Hardware acceleration for AI (FPGA, softcore GPU/FGPU) 3. Reliability of AI systems 4. Approximate Computing 5. Neuromorphic Computation (SNN) 6. Model optimization (pruning, quantization, LLM/transformer optimization) The work focuses on the development and optimization of advanced AI models such as transformers and LLMs, as well as the design of energy-efficient, reliable, and secure hardware accelerators. Research areas include hardware-aware optimization, fault-tolerant architectures, approximate computing methods, neuromorphic computing, and model compression for edge AI and safety-critical applications. Your own topic suggestions are also negotiable. Appointments can only be arranged via email.