Academic Assistant - Computer Science, Mathematics (m/f/d)

Karlsruhe Institute of Technology (KIT) – The University of the Helmholtz Association In dialogue with society, KIT develops solutions for major challenges – from climate change, energy transition and sustainable use of natural resources to artificial intelligence, technological sovereignty and demographic change. As the University in the Helmholtz Association, KIT combines scientific excellence from knowledge gain to application orientation under one roof – and is thus in a unique position to advance this transformation. As a university of excellence, KIT offers its more than 10,000 employees and its 22 800 students outstanding opportunities to shape a sustainable and resilient future. KIT – Science for Impact. We are looking for the Laboratory for Applications of Synchrotron Radiation (LAS) for 3 years Senior Consultant (m/d) - Computer Science, Mathematics System development for data-intensive 3D image reconstruction and analysis The new MorphoSphere project develops an advanced data management architecture for distributed, intelligent and interactive storage and analysis of large 3D tomographic data sets. It integrates large-scale research imaging systems (PETRA III at DESY, KIT Light Source), high-performance computing and data infrastructure at DESY, KIT and Heidelberg University as well as various scientific communities. The aim is to optimize data transfer, processing and visualization and thus to accelerate scientific discoveries in different research areas. The exponential growth of imaging data in large-scale research facilities such as synchrotrons poses considerable challenges in terms of data handling, analysis and suitability. Traditional methods are often unable to efficiently process datasets in the Petabyte area, which limits the extraction of meaningful insights from complex 3D experiments. MorphoSphere meets this challenge by combining distributed computing, data federing and artificial intelligence to enable interactive, scalable and intelligent analysis workflows. This approach reflects the general change towards data-driven research infrastructures that combine high-performance computing with machine learning and promote interdisciplinary cooperation and open data practices.

You develop a modular, flexible and expandable system for large scale 3D image processing (approximately 30 GB input data per dataset, thousands of datasets per week). The system creates image processing pipelines in graph-based structure and ensures that all available resources (GPUs in multi-GPU servers as well as several such servers) are used efficiently. Both classic image processing algorithms and machine learning methods should be able to be integrated into these graphs as nodes. The system is used from various programming languages (C, Python). You also cooperate in implementing the nodes, in particular in algorithms for 3D image reconstruction, including typical pre- and post-processing steps. In addition, you optimize the system for different hardware platforms. Their work allows scientists to quickly design, adapt and execute complex image processing pipelines and thus make advanced image analysis more accessible and efficient. Your tasks in detail: Development and development of a plugin-based system for data-intensive image processing pipelines with cross-technology support (CPU, GPU, multi-node systems) with focus on zero-copy data paths, efficient use of fast interconnects and optimal scalability across multiple GPUs and GPU nodes Integration and implementation of image processing algorithms, in particular in the field of 3D imaging Performance optimization for the computing infrastructures of the partner institutions Implementation of a high-level creation of data processing pipelines in Python as well as a visual programming tool and integration into an interactive online platform

You have a PhD in Computer Science, Mathematics or a related subject or alternatively a university degree (Master/Diplom (Uni)) in one of the above-mentioned fields with at least two years of professional experience. In addition, you will bring with you experience in design and implementation of technology-wide systems (CPU, GPU, Cluster); experience with parallel programming (CUDA/OpenCL); very good programming knowledge, especially in C and Python; safe handling of Linux; Knowledge of image processing, machine learning or network technologies is advantageous.

Part of the only German university of excellence with large-scale national research and work on current research and teaching for our future under excellent working conditions in an international environment. Start professionally with targeted training as well as wide-ranging further training offers. Benefit from flexible working time models (Gleitzeit, Homeoffice), various sports and leisure activities as well as child and holiday care services. We also offer a grant to the BW Job Ticket, amounting to 25 €/month and a versatile offer in different menses. We aim to achieve the most uniform possible occupation of jobs with employees (w/m/d) and would therefore be particularly pleased with women's applications. In the case of the same suitability, people with difficulty are recognised. .