About Me

My Name is Roni Häcki

I'm a PhD student at the Systems Group at ETH Zürich under the supervision of Prof. Timothy Roscoe since December 2015. I graduated from ETH Zürich in 2015 with a MSc in Computer Science with specialisation in Distributed Systems.

I'm part of the Barrelfish team. Barrelfish is a research operating system driven by the two problems of scalability for multi- and many-core systems and heterogeneous system hardware. Barrelfish treats a single machine as a distributed system. In consequence, techniques of distributed systems can be applied. Currently the Barrelfish team is working on extending Barrelfish to rack-scale systems

My current key interest is in how to manage the network as a resource within a rack-scale system. Within a rack-scale system fast communication links blur the line between machines, letting it look different from a traditional network. In a rack-scale system, the information about an applications network resource requirement as well as the underlying hardware is available. This additional information can potentially be incorporate into routing and congestion control decisions.

Publications

2016

Stefan Kaestle, Reto Achermann, Roni Häcki, Moritz Hoffmann, Sabela Ramons, Timothy Roscoe. Machine-aware Atomic Broadcast Trees for Multicores (OSDI 2016) [ pdf ]

Roni Häcki. Replication in Rack-scale Systems (EuroSys Doctoral Workshop 2016) [ pdf ]

Education

BSc Computer Science, ETH Zürich (2010-2013)

Roni Häcki. Dedicating NIC hardware queues to applications on the Barrelfish OS. Bachelors's thesis, ETH Zürich, September 2013
[ pdf ]

MSc Computer Science, ETH Zürich (2013-2015, Specialization: Distributed Systems)

Roni Häcki. Consensus on a multicore machine. Master's thesis, ETH Zürich, September 2015 [ pdf ]

Teaching

System Porgraming and Computer Architecture

Fall 2016. Introduction to computer architecture and system programming:Instruction sets, storage hiearchies, runtime structures with an emphasis on computers as engines for the execution of compiled programs. Interaction between system software and the hardware. Problems that arise from the final respresentation, performance measurement and tuning, and program portability issues are covered. [ course site ][ course catalogue ]

Parallel Programming

Spring 2016. Introduction to parallel programming: deterministic and non-deterministic programs, models for parallel computation, synchronization, communication, and fairness. [ course site ][ course catalogue ]

Contact

Roni Häcki
roni.haecki@inf.ethz.ch
+41 44 632 43 28
Dept. of Computer Science
CAB E 69, ETH Zürich
Universitätstrasse 6
8092 Zürich
Switzerland