Philippe Bonnet

Logo

bonnet@di.ku.dk

Professor

Datalogisk Institut Københavns Universitet (DIKU)

PLTC Section

Universitetsparken 5
2100 Copenhagen, Denmark

ORCID logo Orcid
Linkedin Logo Linkedin
DBLP Logo DBLP
Google Scholar Logo Google Scholar

News

Bio

Philippe Bonnet is professor at DIKU, the computer science department of the University of Copenhagen.

He is an experimental computer scientist with a background in database systems. For thirty years, he has explored the design, implementation, and evaluation of database systems in the context of successive generations of computer classes, including wireless sensor networks, computer clusters and most recently computational storage. Philippe has a track record of successful research projects under DARPA, NSF (while a research associate at Cornell University), EU and Danish funding (at DIKU and ITU).

Philippe is an expert on storage system software. In this area, he contributed to the uFlip Benchmark, the Linux multiqueue block layer, the Linux framework for Open-Channel SSDs, the OX architecture for computational storage, the xNVMe library and Delilah, a prototype for eBPF offload on computational storage. Philippe was involved in the development of sensor networks in the early 2000s with the Cougar project at Cornell, and the Hogthrob and Mana projects at U.Copenhagen. Philippe is co-author of a book on database tuning together with Dennis Shasha.

Philippe joined DIKU in 2001 as assistant professor in the Distlab group. He was faculty at the IT University from 2009 to 2023, and he is now back at DIKU. Before Denmark, Philippe worked as software engineer at ECRC in Munich, Germany from 1993 to 1996, as a PhD student at GIE Dyade in Grenoble, France, and as a postdoc and later research associate at Cornell University, NY, USA from 1999 to 2001. In this period, Philippe had the priviledge to work under the supervision of Stephane Bressan at ECRC, Anthony Tomasic at Dyade and Praveen Seshadri at Cornell.

Philippe is a trustee of the VLDB Endowment and currently chairs the ACM EIG on Reproducibility and Replicability.

PhD Students

Students

Kasper Hjort Berthelsen (ITU, 2024)     Morten Tychsen Clausen (ITU, 2024)    Niclas Hedam (ITU, 2024)    

Graduates

Matias Bjørling (ITU, 2015)     Marcus Chang (DIKU, 2009)     Niv Dayan (ITU, 2015)     Jonathan Furst (ITU, 2016)     Javier Gonzalez (ITU, 2015)     Joel Granados (ITU, 2013)     Aslak Johansen (ITU, 2014)     Martin Leopold (DIKU, 2007)     Ivan Luiz Picoli (ITU, 2019)

Research

Many of the research questions I study are motivated by three recent trends that drive the evolution of computer systems:

  1. The performance of modern storage hardware has triggered a necessary redesign of the storage software stack.
  2. The diversification of workloads driven by artificial intelligence and big data, combined with the limits reached by central processing units has propelled the emergence of specialized hardware to deliver top performance at an acceptable energy cost.
  3. Data movement bottlenecks in traditional, processor-centric architectures have led to the development of data-centric architectures, where processing is moved to where data is stored.

My research is based on the hypothesis that solid-state drives that are programmed and specialized together with database management systems minimize data movement and improve performances. This hypothesis is based on our early work on th performance characteristics of SSDs and on the integration of SSDs in Linux. This work demonstrated the limitations of layered system design with generic SSD firmware. The idea is that databases and SSDs should be co-designed.

Currently, my research focuses on:

Teaching

Fall 2023:

Spring 2024:

Service