network applications and help to transfer them to neuromorphic IC’s and architectures.
What you will do
You take part in the development of the next generation neuromorphic solutions. Your focus is on developing an innovative application prototype using neural network algorithms and combining the achievements from imec technology research (IC’s and sensors) with state-of-art components. Apart from designing embedded prototypes, you conduct research in new neural network architectures in close collaboration with the IC design team. You collect the requirements from various stakeholders and build a software solution. Together with the other neuromorphic researchers you will propose new neuromorphic IC architectures. You will demonstrate these new designs and concepts in compelling demonstrators.
Your responsibilities include:
- Feed the neuromorphic IC designs with innovative implementations and new algorithms based on real applications.
- Develop software prototypes that uses neural algorithms on GPU and on our new neuromorphic IC’s.
- Make system trade-offs for minimal power consumption at best performance and low cost, while keeping an eye on
maintainability and re-usability.
- Test, characterize and evaluate (parts of) the prototypes.
- Demonstrate the prototypes and solutions to our partners and prospects.
- Publish your results in high level journals and conferences.
- Deliver results on time in line with the roadmap.
- Actively contribute to the evolution of our neuromorphic roadmap.
What we do for you
As a Neuromorphic Researcher at imec the Netherlands you get the chance to scale future computing beyond what is possible with silicon CMOS. In collaboration with your colleagues, you will work on a number of challenges to the design of efficient brain-like chips: scalable, energy-efficient, highly flexible network architectures that can learn with incredible efficiency.
Who you are
- PhD in Computer Science or Electronical Engineering.
- 6+ years of experience in neural networks.
- Proven experience in actively building applications and adapting neural algorithms in order to improve performance.
- Hands on experience with neural networks and various simulators.
- Experience in simulating and comparing different real time neural network architectures or digital processor designs is preferred.
- Experience in sensor fusion applications or real time feedback loops is preferred.
- Proven can-do attitude, coming up with practical working solutions is a must.
- Good communication skills in English.