We design hardware systems for artificial intelligence, focusing on efficient, high-performance architectures that meet modern AI applications' growing demands. Our work spans the full stack of hardware-software co-design, bridging algorithm development, system architecture and compilers to deliver intelligent and energy-efficient solutions. While our primary goal is to build better hardware for AI, we also draw inspiration from the brain in several of our projects — using principles such as parallelism, sparsity, and locality to rethink traditional computing models by learning from its efficiency and adaptability. Our goal is to give our members the opportunity to do amazing research and possibly publish their work.
This project is part of the prestigious CAD Contest at ICCAD, focusing on optimizing digital designs for power and timing using multibit flip-flops - an advanced technique in physical design and synthesis. You'll work on developing algorithms that analyze and transform RTL/netlists to meet strict timing and energy constraints. Link to the competition website: https://www.iccad-contest.org/
Recommended skills:
In this project, we explore how to perform AI computations directly in memory using upMEM’s PIM (Processing-In-Memory) DRAM modules. Inspired by the brain's tight integration of memory and computing, this approach helps reduce data movement and boost energy efficiency in AI workloads.
Recommended skills:
This project focuses on integrating a neuromorphic accelerator with the PYNQ platform (Python + Zynq SoC). You'll work at the intersection of hardware acceleration and embedded systems, enabling real-time brain-inspired computation for robotics and edge AI applications.
Recommended skills:
Within this group, we are developing a brain-computer interface (BCI). By collecting EEG signals, processing them, and classifying them, we could offer quadriplegic people the possibility of controlling virtual environments (e.g. mouse cursor on a screen) with only their brains. We envision this to be translated into the real world into a project such as a mind-controlled wheelchair or robotic arm. This team consists of different sub-teams that work together to put the different pieces of the BCI together. The first milestone of this team was the participation in the ETH Cybathlon 2024 - BCI race. We are now continuing our work to improve the system and planning the organization of possible sub-projects to work on in the BCI team for the Summer Semester 2025.
Help design and conduct EEG experiments that aim to enable a person to robustly control a set of controls that can be mapped to, for example, play a computer game.
Nice to have:
Research and experiment on methods to benchmark and test our BCI system.
Nice to have:
Nice to have:
Nice to have:
Research on neural data privacy and security of EEG-based mobile BCI systems.
Nice to have:
Nice to have:
The Communications & Operations team is the backbone of neuroTUM's internal and external affairs. On the external side, we aim to make science accessible to all, showcasing the club's work on social media and organizing events and seminars where people can network with experts in neurotechnology, as well as get a better understanding of our projects and various advancements in the field. We promote scientific communication through newsletters, blog posts and short informational or work session videos. Internally, we strive to enable our project teams to focus solely on their project work by handling legal, administrative and operational tasks in cooperation with the club's leadership, as well as by acquiring sponsors and collaborators, ensuring that our club has all the necessary resources to grow and bloom.
We are looking for:
The Web Development Team Member ensures that neuroTUM’s digital presence reflects our commitments. By building and maintaining our new React-based website, you'll create engaging digital experiences that effectively communicate neuroTUM’s mission and activities.
The role involves developing responsive web components, enhancing user experience, and collaborating closely with interdisciplinary teams to integrate innovative functionalities.
Nice to have:
What we offer:
The Ethics and Responsibility Officer ensures that ethics is not just a theoretical concern but a fundamental part of neuroTUM’s engineering and research processes. By embedding structured ethical engagement, decision-making tools, and user-centred design principles, the team will help establish neuroTUM as a model for responsible student-led innovation in neurotechnology. The role involves organising ethical workshops for the community body (in G&D, neuroethics, AI ethics etc.) curating resources for ethical education of our team, and embedding ethical tools and engagement into neuroTUM’s culture and processes.
Nice to have:
What we offer:
Working on neurotechnology-related or research projects is not a cheap hobby to have. Our teams need financial and material means to achieve their work reliably. As a student organization, we rely on constant support from research chairs and industry partners.
Nice to have (Not necessary):
What we offer:
Our organization needs a good overview over the budget, especially because of the many projects that run in parallel, all in need of money. Moreover, as a registered association (non-profit) we have a legal requirement of conducting thorough accounting.
Nice to have:
What we offer:
As a registered association (non-profit) we have a must to abide by the laws and regulations of Germany. To achieve this, we need to navigate and understand these topics well, how they impact us and how we can implement them.
Nice to have:
What we offer:
Be it project management or computing resources, infrastructure is the bread & butter of a research organization. Without structure, resources and tools that enable our project teams to achieve their goals, it is very hard to push forward. As part of this team you will help set a solid foundation that allows our teams to work efficiently and stress free.
Nice to have:
What we offer:
The goal of this team is the design and build of a custom Electroencephalogram (EEG) system, including active electrodes. This device is a key component of a brain-computer interface (BCI), which allows the collection of neuronal data. As commercial systems are prohibitively expensive despite comparatively low material cost, we have set out to build our own. Within this team, members will get to dive into the world of circuit & PCB design for both analogue and digital systems, soldering, and embedded programming for microcontrollers. Take a look at our social media to see what we've been up to!
Nice to have:
We are developing an algorithm aiming to diagnose children with rare diseases. We plan to train deep learning models using a variety of data sources, including multiomics data, presented in a graph format.
We are looking for:
Tasks:
We are a newborn neuroTUM team seeking to find applications of BCI-achievements in everyday life of disabled people. Our current project is a BCI-controlled robotic arm! We have a few exciting development paths for the future but we will start with the basics such as robotic arm simulations. Please join our team if you are interested in using your software knowledge in bringing neuroengineering methods closer to real-life applications. We will learn and grow together!
.png)
