Conducting research for a changing society: This is what drives us at Forschungszentrum Jülich. As a member of the Helmholtz Association, we aim to tackle the grand societal challenges of our time and conduct research into the possibilities of a digitized society, a climate-friendly energy system, and a resource-efficient economy. Work together with around 6,800 employees in one of Europe’s biggest research centres and help us to shape change!
Join us in making better natural language models for European languages. In the OpenGPT-X project (a GAIA-X project), together with our partners we will improve existing models, applying Machine Learning (ML) techniques at large-scale on the latest GPU-based supercomputers. The optimized models, training code, and optimizations will be published open source to sustainably support and further develop the field.
Jülich Supercomputer Centre (JSC) operates one of the most advanced supercomputing infrastructures for scientific and engineering applications in Europe and provides them to the research community. Amongst them is JUWELS Booster, currently Europe’s fastest supercomputer utilizing 3600 NVIDIA A100 GPUs to achieve more than 70 PFLOP/s performance. The “Accelerating Devices Lab” at JSC offers consulting and helps to optimize applications on the latest hardware acceleration devices at largest scale. For this purpose, we work together with various domain scientists and technology vendors, analyzing how to exploit the massive computer power of current and next generation devices at massive scale.
We are looking to recruit a
Research Scientist – High Performance Computing (HPC) / Machine Learning (ML)
Your Job:
Within the Accelerating Devices Lab, you will work at the intersection of High Performance Computing and Machine Learning, becoming a key enabler of the massive-scale language models needed in OpenGPT-X.
Specifically, you will:
- Run and optimize scalable training algorithms for a large-scale language model in close collaboration with our partners in the OpenGPT-X project
- Analyze, consult on, and improve training runs using up to the full JUWELS Booster machine, exploiting the features of the underlying hardware
- Be the liaison for hardware-specific consulting within the project at JSC
- Identify key code requirements in relation to the use of new hardware for Natural Language Processing and other ML-related workloads
- Present your research at scientific meetings, conferences, and as scientific papers
Your Profile:
You are an excellent team player and a curious researcher, knowledgeable within the field of HPC, especially of GPU-accelerated HPC systems. You worked with and tuned ML / DL workloads before. You know how to dig into complex applications and scale them to 11. You are interested in working with state-of-the-art supercomputers, to define applications and platforms for the next generation of supercomputers.
More specifically, your profile is:
- Excellent master’s degree or PhD (preferred) in Computer Science, Mathematics, Physics, or a similar fields
- Comprehensive experience using HPC Systems as well as parallel/distributed programming, including the usual tools and programming / scripting languages
- Practical experience in optimizing AI and ML workloads, preferably with special emphasis on NLP models
- A track record of your work in code repositories (open source), scientific workshop, conferences, and other publications
- Very good command of written and spoken English
- Self-motivated personality, used to work in a multidisciplinary team and environment solving scientifically challenging problems on the largest computers in the world
Our Offer:
We work on the very latest issues that impact our society and are offering you the chance to actively help in shaping the change! We support you in your work with:
- An exciting and varied role at the frontiers of science in an international and interdisciplinary working environment, with a possibility to shape the next generation of HPC systems
- Access to cutting-edge and unique supercomputing systems including large-scale GPU installations and Quantum Computers
- Comprehensive training courses and individual opportunities for personal and professional further development
- Extensive company health management
- Ideal conditions for balancing work and private life, as well as a family-friendly corporate policy
- Full-time position with the option of slightly reduced working hours and 30 days of annual leave
- Targeted services for international employees, e.g. through our International Advisory Service
- A large research campus with green spaces, offering the best possible means for networking with colleagues and pursuing sports alongside work
We offer you an exciting and varied role in an international and interdisciplinary working environment. The position is initially for a fixed term of three years, with possible long-term prospects. Salary and social benefits are in conformity with the provisions of the Collective Agreement for the Public Service (TVöD).
Please note a similar open position at JSC within OpenGPT-X focusing less on the hardware-related aspects but more on the algorithmic parts of large-scale ML in HPC.
Forschungszentrum Jülich promotes equal opportunities and diversity in its employment relations. We also welcome applications from disabled persons.
We look forward to receiving your application until October 24, 2021 via our Online Recruitment System!
Questions about the vacancy?
Get in touch with us by using our contact form. Please note that for technical reasons we cannot accept applications via email.