• Hubert Eichner

  • GMail: hubert.georg.eichner
  • Seattle, WA
  • 2001 - 2007: CS student at TU Munich
  • 2008 - 2012: PhD student at Max-Planck-Institute of Neurobiology
  • 2012 - 2015: Senior Scientist at Microsoft, Bellevue WA
  • since 2015: Senior Staff Software Engineer at Google, Seattle WA
  • internships, summer courses, student jobs:
    • 2004: Transaction Software GmbH
    • 2004 - 2006: admin at TUM's InfiniBand cluster
    • 2005: IBM Extreme Blue, Logic Design
    • 2009: student @ Methods in Computational Neuroscience, Woods Hole, MA
    • 2011: teaching assistant @ MCN, Woods Hole, MA
  • interests:
    • machine learning
    • physics simulations (game physics, neural simulations, ...)
    • computer vision
    • stuff I can't figure out
  • theses & some publications:
    • Internal Structure of the Fly Elementare Motion Detector (2012). PhD thesis
    • Parallel Neural Simulations on Multi-Core Processsors (2007). Diploma thesis
    • Wang et al. A Field Guide to Federated Optimization (2021). Work done at Google.
    • Kairouz et al. Advances and Open Problems in Federated Learning (2021). Foundations and Trends in Machine Learning, 14(1-2)., section "Addressing System Challenges". Work done at Google.
    • Wang, Mathews, Kiddon, Eichner, Beaufays, Ramage. Federated Evaluation of On-Device Personalization (2019). Work done at Google.
    • Eichner, Koren, McMahan, Srebro, Talwar. Semi-Cyclic Stochastic Gradient Descent (2019). Work done at Google.
    • Bonawitz, Eichner, Grieskamp, Huba, Ingerman, Ivanov, Kiddon, Konecny, Mazzocchi, McMahan, Van Overveldt, Petrou, Ramage, Roselander. Towards Federated Learning at Scale: System Design (2019). Work done at Google.
    • Yang, Andrew, Eichner, Sun, Li, Kong, Ramage, Beaufays. Applied Federated Learning: Improving Google Keyboard Query Suggestions (2018). Work done at Google.
    • Hard, Rao, Mathews, Beaufays, Augenstein, Eichner, Kiddon, Ramage. Federated Learning for Mobile Keyboard Prediction (2018). Work done at Google.
    • Federated Learning: Collaborative Machine Learning without Centralized Training Data (2017). Blog post and work done at Google.
    • Joesch, Weber, Eichner, Borst. Functional specialization of parallel motion detection circuits in the fly (2013). J Neurosci, 33
    • Eichner, Joesch, Schnell, Reiff, Borst. Internal Structure of the Fly Elementare Motion Detector (2011). NEURON, 70(6)
    • Eichner, Borst. Hands-On Parameter Search for Neural Simulations by a MIDI-Controller (2011). PLoS ONE, 6(10)
    • Spavieri, Eichner, Borst. Coding Efficiency of Fly Motion Processing is Set by Firing Rate, Not Firing Precision (2010). PLoS Computational Biology, 6(7)
    • Eichner, Klug, Borst. Neural Simulations on Multi-Core Architectures (2009). Frontiers in Neuroinformatics, 3
    • Hines, Eichner, Schuermann. Neuron Splitting in Compute-Bound Parallel Network Simulations Enables Runtime Scaling With Twice as Many Processors (2008). Journal of Computational Neuroscience, 25
    • Weber, Eichner, Cuntz, Borst. Eigenanalysis of a Neural Network for Optic Flow Processing (2008). New Journal of Physics, 10
    • Stodden, Eichner, Walter, Trinitis. Hardware Instruction Counting for Log-Based Rollback Recovery on x86-Family Processors. Service Availability, Lecture Notes in Computer Science, 4328
    • Eichner, Trinitis, Klug. Implementation of a DSM-System on Top of InfiniBand. Special Session on "Parallel and Distributed Storage Systems" at 1th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP2006), Montbeliard-Sochaux, France