- Clemens Lutz, PhD

Developer Technology Engineer at NVIDIA.

My professional interests are data management using modern hardware and distributed systems. Currently I am investigating how we can use fast, next-generation interconnects such as NVLink to scale data management on GPUs.

Previously, I successfully achieved my PhD in Computer Science at TU Berlin, where I was mentored by Volker Markl and cooperated with Tilmann Rabl, Sebastian Breß, and Steffen Zeuch.

Before starting my PhD studies, I received my MSc degree from ETH Zurich in collaboration with IBM Research, Zurich under supervision of Thomas R. Gross and Animesh Trivedi. My Master's was preceded by an exchange at Imperial College London, where I wrote my BSc thesis under supervision of Peter Pietzuch and Paolo Costa, during my Bachelor's studies at ETH Zurich.

clutz@nvidia.com

NVIDIA
2788 San Tomas Expy
Santa Clara
CA 95051
U.S.A.

News

17 March 2025: I'll be at the NVIDIA GTC connect with experts session on Accelerating Data Analytics Workflows on GPU Systems on Tuesday. Feel free to drop by!
16 March 2025: Our paper on Efficiently Indexing Large Data on GPUs with Fast Interconnects has been published at EDBT'25. My congratulations to Josef Schmeißer as first author!
28 February 2025: The German Informatics Society has announced that they will grant me the DBIS Dissertation Award. I'm deeply humbled by this recognition of my work. Next week at BTW 2025, I will present a short contribution that summarizes my dissertation. Looking foward to BTW!
26 January 2025: Query processing on heterogeneous hardware is a big topic. That's why we are providing a background in our new book chapter, available to you with open access. The book is titled "Scalable Data Management for Future Hardware" and includes many insights obtained from database research. Take a look!
20 April 2024: I'm honored to announce that the ICDE program chairs have awarded me Outstanding Reviewer for ICDE 2024!
8 April 2024: Heads-up, I've relocated to NVIDIA HQ in Santa Clara. It's a great opportunity to work in-person with the DevTech Compute team members located here. Looking forward to avoiding the 9 hour time zone shift for meetings!

Publications

EDBT'25 Paper Code: Efficiently Indexing Large Data on GPUs with Fast Interconnects Josef Schmeißer, Clemens Lutz, Volker Markl, in the 28th International Conference on Extending Database Technology, March 25–28, 2025, Barcelona, Spain.
BTW'25 Short Paper: Scalable Data Management using GPUs with Fast Interconnects — A Brief Overview Clemens Lutz, in Database Systems for Business, Technology and Web, March 3–7, 2025, Bamberg, Germany.
Springer Book Chapter: Query Processing on Heterogeneous Hardware Anastasiia Kozar, Janis von Bleichert, Sebastian Breß, Philipp M. Grulich, Clemens Lutz, Tilmann Rabl, Viktor Rosenfeld, Jonas Traub, Steffen Zeuch, Volker Markl, book chapter, Springer, Cham, Switzerland, January 24, 2025.
EDBT'24 Paper Code: Benchmarking Stream Join Algorithms on GPUs: A Framework and its Application to the State of the Art Dwi P. A. Nugroho, Philipp M. Grulich, Steffen Zeuch, Clemens Lutz, Stefano Bortoli, Volker Markl, in the 27th International Conference on Extending Database Technology, March 25–28, 2024, Paestum, Italy.
DBIS Dissertation Award PhD 2022 Thesis Slides: Scalable Data Management using GPUs with Fast Interconnects. Clemens Lutz, PhD thesis, Faculty IV, TU Berlin, Berlin, Germany, November 2022.
SIGMOD'22 Paper Code YouTube Slides Poster: Triton Join: Efficiently Scaling to a Large Join State on GPUs with Fast Interconnects Clemens Lutz, Sebastian Breß, Steffen Zeuch, Tilmann Rabl, Volker Markl, in the ACM International Conference on Management of Data, June 12–17, 2022, Philadelphia, PA, USA.
DaMoN'21 Paper Code Slides: An Energy-Efficient Stream Join for the Internet of Things Adrian Michalke, Philipp M. Grulich, Clemens Lutz, Steffen Zeuch, Volker Markl, in the 17th ACM Int. Workshop on Data Management on New Hardware (DaMoN'21), held online with SIGMOD/PODS, June 21st, 2021.
BTW'21 Best Paper Reproducible Paper Slides: Fast CSV Loading Using GPUs and RDMA for In-Memory Data Processing Alexander Kumaigorodski, Clemens Lutz, Volker Markl, in Database Systems for Business, Technology and Web, September 13–17, 2021, Dresden, Germany.
SIGMOD'20 Best Paper Reproducible Paper Blog YouTube Slides: Pump Up the Volume: Processing Large Data on GPUs with Fast Interconnects Clemens Lutz, Sebastian Breß, Steffen Zeuch, Tilmann Rabl, Volker Markl, in the ACM International Conference on Management of Data, June 14–19, 2020, Portland, OR, USA.
PVLDB'19 Paper Code: Analyzing Efficient Stream Processing on Modern Hardware Steffen Zeuch, Bonaventura Del Monte, Jeyhun Karimov, Clemens Lutz, Manuel Renz, Jonas Traub, Sebastian Breß, Tilmann Rabl, Volker Markl, in The Proceedings of the VLDB Endowment, Vol. 12, No. 5, Los Angeles, CA, USA, August 26th-30th, 2019.
Datenbanken Spektrum Paper Blog Code: Efficient and Scalable k-Means on GPUs Clemens Lutz, Sebastian Breß, Tilmann Rabl, Steffen Zeuch, Volker Markl, in Datenbanken Spektrum 2018.
DaMoN'18 Paper Code Slides Poster: Efficient k-Means on GPUs Clemens Lutz, Sebastian Breß, Tilmann Rabl, Steffen Zeuch, Volker Markl, in the 14th ACM Int. Workshop on Data Management on New Hardware (DaMoN'18), colocated with SIGMOD/PODS, Houston, TX, USA, June 11th, 2018.
ICDCS'15 Paper: RStore: A Direct-Access DRAM-based Data Store Animesh Trivedi, Patrick Stuedi, Bernard Metzler, Clemens Lutz, Martin Schmatz, Thomas R. Gross, in the 35th IEEE Int. Conf. Distributed Computing Systems (ICDCS'15), Columbus, OH, USA, June 29th - July 2nd, 2015.
MSc 2014 Thesis: Carafe: High-Performance, In-Memory Graph Processing with RDMA. Clemens Lutz, MSc thesis, D-INFK, ETH Zurich, Zurich, Switzerland, October 2014.

Service

Theses & Teaching

Thesis Supervision

Thesis: Cooperative Heterogeneous Query Execution. Apostolos Planas, MSc thesis, Faculty IV (EECS), TU Berlin, Berlin, Germany, February 2022
Thesis: Fast CSV Loading Using GPUs and RDMA for In-Memory Data Processing. Alexander Kumaigorodski, MSc thesis, Faculty IV (EECS), TU Berlin, Berlin, Germany, July 2020
Thesis: Lock-based Data Structures on GPUs with Independent Thread Scheduling. Phillip Grote, BSc thesis, Faculty IV (EECS), TU Berlin, Berlin, Germany, February 2020

Teaching

In-Memory Databases On Modern Hardware, SS 2021, TU Berlin
Database Systems Seminar, SS 2021, TU Berlin
Big Data Analytics Seminar, WS 2020, TU Berlin
Big Data Analytics Project, WS 2020, TU Berlin
Information Systems and Data Analysis, SS 2020, TU Berlin
In-Memory Databases On Modern Hardware, SS 2020, TU Berlin
Databases Laboratory, WS 2019, TU Berlin
Information Systems and Data Analysis, SS 2019, TU Berlin
In-Memory Databases On Modern Hardware, SS 2019, TU Berlin
Databases Laboratory, WS 2018, TU Berlin
Information Systems and Data Analysis, SS 2018, TU Berlin
Databases Laboratory, WS 2017, TU Berlin
Information Systems and Data Analysis, SS 2017, TU Berlin
Databases Laboratory, WS 2016, TU Berlin
Information Management Seminar, SS 2016, TU Berlin
Databases Seminar, WS 2015, TU Berlin
Operating Systems and Networks, SS 2011, ETH Zurich