Throughout this project, we would like to design benchmarking methodology and apply it on a few advanced neuromorphic processors by performing reproducible power and performance measurements for some specific applications.
What you will do
To be considered for this position: The European candidates must be enrolled in a university or university of applied science. Non-European master students who are enrolled in a Dutch university are also welcomed to apply.
In the neuromorphic group of Imec (Holst-Centre), we are designing neuromorphic processors to execute scalable Edge AI applications. One important activity in this process is to benchmark our product and compare it with other available neuromorphic processors.
A proper benchmark for neuromorphic engines is still an open problem . In this project, you will design a benchmarking methodology. As part of designing this benchmarking methodology you will also choose, together with us, an application which is good for the benchmarking. You will implement the application on one of the two neuromorphic processors which are available for us in imec (Epiphany processor , SpiNNaker ) and apply the benchmarking methodology. Optionally we would like to benchmark also our own neuromorphic processor with the same methodology and application.
You will work in imec Holst Centre and will contribute to the European TEMPO project (Technology and hardware for neuromorphic computing). The outcome of the project may be published in high-impact conferences. The interested applicants should submit their CV, the academic transcripts (including the scores and the courses), and (if known) the name of the project supervisor from the university.
 Benchmarks for progress in neuromorphic computing (https://www.nature.com/articles/s42256-019-0097-1)
 Epiphany processor (https://www.parallella.org/)
 SpiNNaker processor (https://en.wikipedia.org/wiki/SpiNNaker)
- Pre-study to become familiar with the applications and architecture of neuromorphic processors.
- Design and implementation of the benchmarking methodologies.
- Measurements of key performance indexes (power, performance, etc.).
- Thesis writing and documentation in Imec.
- And if the results are promising a publication of the results.
Who you are
- Master Student in Electrical/Computer engineering.
- Available for 6 or 9 months.
- Very good in embedded C (++) programming.
- Knowledge of neural network and neuromorphic architectures are a plus.
- Good written and verbal English skills.
Click on 'apply' to submit your application. You will then be redirected to e-recruiting.