HZB Newsroom

With increasing size and complexity, quantum computers become a sort of black box. Using methods from mathematical physics, a team has now succeeded in deriving concrete numbers from random, data sequences that can serve as a benchmark for the performance of a quantum computer system. Experts from Helmholtz-Zentrum Berlin, Freie Universität Berlin, Qusoft Research Centre Amsterdam, the University of Copenhagen and the Technology Innovation Institute Abu Dhabi were involved in the work, which has now been published in Nature Communications.

X-ray microscopy (cryo-SXT) enables high-resolution insights into cells and cell organelles - in three dimensions. Until now, the 3D data sets have been analysed manually, which is very time-consuming. A team from Freie Universität Berlin has now developed a self-learning algorithm based on a convolutional neural network. In collaboration with experts in cell biology (FU Berlin) and X-ray microscopy at the Helmholtz Zentrum Berlin, this algorithm has now been used for the first time to analyse cell components in cryo-SXT data sets. It identified cell organelles and produced highly detailed, complex 3D images within a few minutes.

Prof. Dr. Jens Eisert heads the Quantum Computation and Simulation research group, which is jointly funded by Freie Universität Berlin and HZB. The theoretical physicist recently received an Advanced Grant from the European Research Council ERC for his research project "DebuQC". But Eisert is not only an award-winning scientist, he is also concerned about global warming. Last year, on the initiative of Ryan Sweke, he and his team published a proposal that deserves attention: scientific publications on theoretical physics or chemistry could include a simple table listing the greenhouse gases emitted during the research. This would raise awareness of the fact that research is not climate neutral.

Are you a doctoral researcher or Postdoc and your research has a strong link to (applied) data and information sciences?

Experimental data is often not only highly dimensional, but also noisy and full of artefacts. This makes it difficult to interpret the data. Now a team at HZB has designed software that uses self-learning neural networks to compress the data in a smart way and reconstruct a low-noise version in the next step. This enables to recognise correlations that would otherwise not be discernible. The software has now been successfully used in photon diagnostics at the FLASH free electron laser at DESY. But it is suitable for very different applications in science.

At HZB, a laboratory for automated X-ray tomography on solid-state batteries is being set up. The special feature: 3D data during charge/discharge processes (operando) can be evaluated quickly and in a more versatile way using artificial intelligence (AI) methods. The Federal Ministry of Research and Education is funding the "TomoFestBattLab" project with 1.86 million euros.

With conventional methods, it is extremely time-consuming to calculate the spectral fingerprint of larger molecules. But this is a prerequisite for correctly interpreting experimentally obtained data. Now, a team at HZB has achieved very good results in significantly less time using self-learning graphical neural networks.

An international team of experts has collected data on metal halide perovskite solar cells from more than 15,000 publications and developed a database with visualisation options and analysis tools. The database is open source and provides an overview of the rapidly growing knowledge as well as the open questions in this exciting class of materials. The study was initiated by HZB scientist Dr. Eva Unger and implemented and coordinated by her postdoc Jesper Jacobsson.