Looking for a researcher/PhD student in energy-efficient machine learning hardware#
To apply for this position, click on the
Apply for this position link at the bottom of the page. Please do not spend more than one hour preparing your responses to the prompts in the application form.
Our university has also posted the same job offer on their job board. Please apply using both my homepage and the university’s page.
I am continuously reviewing the applications. Applications are welcome until the position is filled.
Q&A Webinar: was on November 8th. The webinar recording is here. I made a mistake regarding compensation: There was a new collective negotiation and the monthly salary in hand will be ~€2800 in hand beginning 1st of March.
I invite anyone who is interested to apply and will assess applications primarily based on anonymized prompt answers so please ensure they represent your fit for the position well. I will put minimal weight on CV and references. Please do not include a cover letter, photograph of yourself or any personal information that is not relevant, e.g., gender, marital status, age, identity traits, etc.
Employment type: limited from Jan 2024 - Dec 2026, ~40h per week, down to 50% part time possible
Location: Deggendorf or fully remote upon agreement. You must be living in Germany.
Email if you have any questions.
About the position#
I am seeking a researcher/PhD student to investigate energy-efficient and fault-tolerant machine learning (ML) hardware as part of a research project supported by a grant. As an ML hardware researcher you will collaborate with me on novel research & development related to energy-efficient training of reinforcement learning workloads, fault-tolerance of ML hardware and impact of emerging hardware architectures on ML. Your work will be crucial for improving the carbon footprint and reliability of the future of AI and its impacts on society.
Your day-to-day activities will be researching the latest developments in the field, discussing new ideas for improvement, evaluating these ideas by conducting experiments on hardware and publishing your findings. Over the course of a year, I anticipate you will have produced 1-2 leading papers. You do not have any teaching duties.
An ideal candidate would have experience in implementing reinforcement learning from scratch on an FPGA, intellectual curiosity and be skilled at communicating technical topics clearly. A Masters degree is necessary. You are (also) encouraged to apply if you have related experience and interest in this field, e.g., you only worked with FPGAs, but do not have any experience with machine learning.
As an ML hardware researcher, your main responsibilities will entail:
Optimizing reinforcement learning algorithms on FPGAs and heterogeneous architectures
Evaluating ideas by implementing hardware demonstrators
Publishing research findings
Staying informed of industry trends and emerging technologies in ML hardware
Being the go-to expert for staff and stakeholders’ queries in the field
What I am looking for#
Intellectual curiosity, open-mindedness
Clear communication skills on technical topics
Competitive advantage compared to me or other candidates in at least one of:
Understanding of foundational reinforcement learning algorithms
Experience with hardware design languages (HDL) or high-level synthesis (HLS)
Knowledge of computing architectures, e.g., GPU, FPGA
What I offer#
search for Compensation above
I can imagine that you can earn much more in the industry as an engineer. I also earned more when I worked in the industry. Please read other benefits before you skip.
AMD Alveo U280 data center accelerator card on a 24-core, 1TB RAM server
Access to other FPGA- and GPU-based accelerators through AMD heterogeneous accelerated computer cluster
AMD Zynq-based boards to be used as an embedded evaluation platform
Depending on what is important for you:
Is meaning important for you?: You will be working on problems which matter for the future and you will be able to share your results with everyone.
Is money important for you?: If you pursue a PhD, you can make more money in the industry after your PhD especially in large companies where compensation may be dependent on your academic qualification.
Flexible work hours
At least 30 days annual leave, more days possible upon agreement with me
Inclusivity and fairness: I would like to have a diverse team and find best people for my team, so please do not hesitate to apply regardless of your age, gender identity/expression, political identity, personal preferences, physical abilities, neurodiversity, or any other background.
Language: Please submit all of your application materials in English and note that I require professional level English proficiency.
Travel: In most cases it is not necessary for this position. However researchers often travel for conferences to present their results and network with specialists in their field.
Accessibility: I am committed to run an inclusive and accessible application process. Please reach out to me for any requests, e.g., chat box use during interviews.
Initial skills assessment and creating a shortlist of people to interview
Interview including a presentation of a work trial
Hiring steps may be repeated in case an appropriate candidate is not found. I will inform you about updates, e.g., if I introduce a new review round and if you are not shortlisted, etc.