Vojtech Aschenbrenner

vojtech.aschenbrenner@epfl.ch

Now

• Building (networked and storage) systems with extreme performance.
  Check DIS/LSVD, BUSE and BS3; now focused on scalable storage stack for heterogeneous HW.
• Teaching Operating Systems at EPFL as a Head TA.
• Advised by Prof. Sanidhya Kashyap in RS3Lab@EPFL.
• Computer science PhD student at EPFL since 2022.
  I received EDIC fellowship. Thank you EDIC!
• Trying to improve Czech academia by writing why and how do PhD abroad and heavily contributing to CSrankings.

Before

• Was a lecturer at Charles University (Dept. of Distributed and Dependable Systems).
  I tried to change the teaching style; systems should be learned by doing.
• Had a business on building customized servers and open-source software deployment.
  I replaced dozens of commercial hypervisors and storage solutions with open-source.
• Was a systems engineer (storage and networking). Also interned at Red Hat twice.
• Was thinking about being an AI researcher in my previous life. I converted to systems.
• Was extremely lucky to work with professors
  • Peter Desnoyers from Northeastern University,
  • Petr Tuma from Charles University and
  • Ondrej Kuzelka from Czech Technical University in Prague.
• Obtained degrees with miscellaneous honors from
  • Charles University (MS in Systems),
  • Czech Technical University in Prague (MS in AI/ML) and
  • Brno University of Technology (BS in general CS).

Research Interests

• Operating Systems, Storage, Networking and Computer Architecture with a particular focus on high-performance I/O.
• Artificial Intelligence and Machine Learning in the previous life.

Professional Activities

• EUROSYS 2023 AEC
• ATC 2022 AEC
• MICRO 2022 AEC
• OSDI 2022 AEC

Publications

[Google Scholar] [GitHub]

Conferences and Journals

• Submission under review.
  USENIX Symposium on Operating Systems Design and Implementation (OSDI 2023)

• Beating the I/O bottleneck: A case for log-structured virtual disks
  Mohammad Hossein Hajkazemi*, Vojtech Aschenbrenner*, Mania Abdi, Emine Ugur Kaynar, Amin Mossayebzadeh, Orran Krieger, Peter Desnoyers (* Joint first authors)
  ACM European Conference on Computer Systems (EUROSYS 2022)
  pdf · code

• LSDN — Manage complex (virtual) networks in cloud environment with Linux kernel facilities
  Vojtech Aschenbrenner, Roman Kapl, Jan Matejek, Adam Vyskovsky
  Technical Conference on Linux Networking (NETDEV 2018)
  pdf · code · slides · talk

• Lifted relational neural networks: Efficient learning of latent relational structures
  Gustav Sourek, Vojtech Aschenbrenner, Filip Zelezny, Steven Schockaert, Ondrej Kuzelka
  Journal of Artificial Intelligence Research (JAIR 2018)
  pdf · code

Workshops

• Lifted relational neural networks
  Gustav Sourek, Vojtech Aschenbrenner, Filip Zelezny, Ondrej Kuzelka
  NIPS Workshop on Cognitive Computation: Integrating Neural and Symbolic Approaches (COCO@NIPS 2015)
  pdf · code

• Horn-clause neural networks
  Vojtech Aschenbrenner, Ondrej Kuzelka
  IEEE ICSC Workshop on Semantic Machine Learning (SML 2014)
  pdf · code

Theses

• BUSE: Block Device in Userspace
  Master Thesis, Charles University
  Best software thesis award
  Real world impact in academia and industry
  pdf · code (buse) · code (bs3)

• Deep Relational Learning with Predicate Invention
  Master Thesis, Czech Technical University in Prague
  Published in JAIR
  pdf · code

• Automated Tool-Chain Testing System
  Bachelor Thesis, Brno University of Technology
  Dean’s price for excellent thesis

Teaching

EPFL

Charles University