Monday, 1 July 2019

In case the links below do not work, you can download a zip file containing all DAMON'19 papers.

10:00 - 11:00 | Session 1

10:00 - 10:15 Welcome

Thomas Neumann (TU Munich and Ken Salem (U Waterloo)

10:15 - 10:30 Cold Storage Data Archives: More Than Just a Bunch of Tapes (Bunjamin Memishi, German Aerospace Center; Raja Appuswamy, Eurecom; Marcus Paradies, German Aerospace Center)

10:30 - 10:45 Improving CPU I/O Performance via SSD Controller FTL Support for Batched Writes (Jaeyoung Do, Microsoft; David Lomet, Microsoft Research; Ivan Luiz Picoli, ITU Copenhagen)

10:45 - 11:00 Event Stream Processing on Heterogeneous System Architecture (Michael Körber, University of Marburg; Jakob Eckstein, University of Marburg; Nikolaus Glombiewski, University of Marburg; Bernhard Seeger, University of Marburg)

11:00 - 11:30 | Coffee Break

11:30 - 13:00 | Session 2

11:30 - 12:30 Keynote Presentation

Performance Scaling with Innovative Compute Architectures and FPGAs (Michaela Blott, Xilinx Research) [slides (PDF)]

12:30 - 13:00 Lightning Talks (5 minutes each)

• Persistent Buffer Management with Optimistic Consistency (Lucas Lersch, TU Dresden; Wolfgang Lehner, TU Dresden; Ismail Oukid, SAP SE)
• Exploiting Intel Optane SSD for Microsoft SQL Server (Kan Wu, UW-Madison; Andrea Arpaci-Dusseau, UW-Madison; Remzi Arpaci-Dusseau, UW-Madison; Rathijit Sen, Microsoft; Kwanghyun Park, Microsoft)
• Freezing Frozen Pages with Multi-Stream SSDs (Hyunwoo Park, SungKyunKwan University; Soyee Choi, SungKyunKwan University; Mijin An, Sungkyunkwan University; Sang Won Lee, Sungkyunkwan University)
• LSM Management on Computational Storage (Philippe Bonnet, IT Univ Copenhagen; Ivan Luiz Picoli, ITU Copenhagen; Pinar Tozun, ITU)
• IPA-IDX: In-Place Appends for B-Tree Indices (Ilia Petrov, Reutlingen University; Sergej Hardock, TU Darmstadt; Andreas Koch, TU Darmstadt; Tobias Vincon, Reutlingen University)

13:00 - 14:30 | Lunch and Poster Session

14:30 - 16:30 | Session 3

14:30 - 15:00 Fresh Thinking Talk

Dark silicon — a currency we do not control (Holger Pirk, Imperial College)

15:00 - 15:15 Fast & Strong: The Case of Compressed String Dictionaries on Modern CPUs (Robert Lasch, SAP SE; Ismail Oukid, SAP SE; Roman Dementiev, Intel Deutschland GmbH; Norman May, SAP SE; Suleyman Demirsoy, Intel Corp.; Kai-Uwe Sattler, TU Ilmenau)

15:15 - 15:30 Exact Set Similarity Joins for Large Datasets in the GPGPU paradigm (Christos Bellas, Aristotle University of Thessaloniki; Anastastios Gounaris, Aristotle University of Thessaloniki)

15:30 - 15:45 GPU-Accelerated Similarity Self-Join for Multi-Dimensional Data (Michael Gowanlock, Northern Arizona University; Ben Karsin, Université libre de Bruxelles)

15:45 - 16:00 Accelerating the Unacceleratable: Hybrid CPU/GPU Algorithms for Memory-Bound Database Primitives (Michael Gowanlock, Northern Arizona University; Ben Karsin, Université libre de Bruxelles; Zane Fink, Northern Arizona University; Jordan Wright, Northern Arizona University)

16:00 - 16:15 Performance Analysis and Automatic Tuning of Hash Aggregation on GPUs (Viktor Rosenfeld, DFKI GmbH; Sebastian Breß, DFKI GmbH; Steffen Zeuch, DFKI Berlin; Tilmann Rabl, Technische Universität Berlin; Volker Markl, Techniche Universität Berlin)

16:15 - 16:30 Fluid Co-processing: GPU Bloom-filters for CPU Joins (Tim Gubner, CWI; Diego Tomé, CWI; Harald Lang, TU Munich; Peter Boncz, Centrum Wiskunde & Informatica)

16:30 - 17:00 | Coffee Break

17:00 - 18:50 | Session 4

17:00 - 17:30 Fresh Thinking Talk

Building real database systems on real persistent memory (Tianzheng Wang, Simon Fraser University)

17:30 - 17:45 Lightning Talks (5 minutes each)

• Relational Queries with a Tensor Processing Unit (Pedro Holanda, CWI; Hannes Mühleisen, CWI)
• In situ Statistics Generation within partially reconfigurable Hardware Accelerators for Query Processing (Andreas Becher, Friedrich-Alexander-Universität Erlangen-Nürnberg; Jürgen Teich, Friedrich-Alexander-Universität Erlangen-Nürnberg)
• On the Impact of Memory Allocation on High-Performance Query Processing (Dominik Durner, TUM; Viktor Leis, Friedrich Schiller University Jena; Thomas Neumann, TUM)

17:45 - 18:00 Towards Practical Vectorized Analytical Query Engines (Orestis Polychroniou, Amazon; Kenneth A Ross, Columbia University)

18:00 - 18:15 Data Compression in NVRAM-centric Data Stores - Comparative Experimental Analysis to DRAM (Mikhail Zarubin, TU Dresden; Patrick Damme, Technische Universität Dresden; Thomas Kissinger, TU Dresden; Dirk Habich, TU Dresden; Thomas Willhalm, Intel Deutschland GmbH; Wolfgang Lehner, TU Dresden)

18:15 - 18:30 Persistent Memory I/O Primitives (Alexander van Renen, TUM; Lukas Vogel, TUM; Viktor Leis, Friedrich Schiller University Jena; Thomas Neumann, TUM; Alfons Kemper, TUM)

18:30 - 18:45 Bridging the latency gap between NVM and DRAM for latency-bound operations (Georgios Psaropoulos,EPFL; Ismail Oukid, SAP SE; Thomas Legler, SAP SE; Norman May, SAP SE; Anastasia Ailamaki, EPFL)

18:45 - 18:50 Workshop Closing

Thomas Neumann (TU Munich) and Ken Salem (U Waterloo)