Master Thesis “Charge-redistribution based in-memory computing for neuromorphic systems ”

Master Thesis “Charge-redistribution based in-memory computing for neuromorphic systems”

Job-ID: 5054/22 | Department: Circuit Design | Limitation: 6 month with option of extension |Earliest Entry Date: 1.7.2022

We offer the possibility to work parallel as a student or research assistant, with a working time of max. 19h per week, remuneration according to the guidelines of the state of Brandenburg on the working conditions of research and student assistants.

Master thesis project: 

In-Memory computing targets high performance low power computing and is an emerging field of research. The goal of this project is to analyze and implement a vector-matrix multiplier using a technology node independent concept, evaluate its accuracy and speed limitations.

The Research/Position:

• Analyzing and evaluation of charge-redistribution bassed in-memory-computing concepts
• Implementation of a matrix-vector multiplier in IHP design environment
• Evaluation the accuracy and speed limitations of the chosen concept
• Potentially physical realization of a test chip
• After successful completion of the master thesis project, a PhD degree can be pursued and will be encouraged

Your Qualifications: 

• Bachelo's degree in the field of electrical engineering
• Good knowledge of analog and digital circuits
• Fundamental hands-on experience in the design and simulation of integrated circuits using CAD tools (Cadence Virtuoso, Spectre) would be desirable
• Very good English language skills

Further Information  Online Application Form