Munich, Germany – 5 October 2026 – Innatera, the leader in neuromorphic compute, and Munich-based audio innovation studio Sofilab are joining forces for the 4th Munich Neuromorphic Hackathon. Their joint challenge, “Neuromorphic Singing Voice Recognition,” is one of eight set by research and industry partners. Student teams will explore ways to recognise singing voices and act as a coach for the singer using neuromorphic computing, with the help of Innatera’s Pulsar evaluation kits.
“Neuromorphic architecture opens up new ways to process signals as they are sampled from sensors, but reaching its potential depends on engineers being able to work with it directly,” says Petruț Bogdan, Neuromorphic Architect at Innatera. “Giving students access to Pulsar and a real audio challenge creates room to experiment, test ideas, and discover approaches we might not have anticipated. That accessibility is how new architectures become a source of innovation.”
The challenge brings together Innatera’s neuromorphic processors and developer tools with Sofilab’s singing voice recognition use case, drawing on its experience in music and interactive audio. Together, the companies are giving participants a problem grounded in the complexities of the human voice, alongside the tools and guidance to explore their own solutions.
Hosted by neuroTUM and fortiss, the 4th Munich Neuromorphic Hackathon runs 5–9 October 2026 at fortiss headquarters. Student teams will work on eight challenges set by research and industry partners, then present their results on the final day.
Explore neuromorphic development with Innatera. Learn about Pulsar, evaluation kits and the developer resources available to help engineers build applications for the sensor edge at www.innatera.com.
