Responsive image

Marios Kogias

Assistant Professor (Lecturer)
Department of Computing
Imperial College London
m.kogias@imperial.ac.uk
Room 407C Huxley
CV

About me

I'm an Assistant Professor (Lecturer in the UK system) in the Department of Computing at Imperial College London and a visiting researcher in the Confidential Computing Group at Microsoft Research Cambridge. I work at the intersection of operating systems, networking, and distributed systems and build systems that can provide tail-tolerance, fault-tolerance, and confidentiality guarantees leveraging emerging datacenter hardware.

Before joining imperial was a researcher at Microsoft Research Cambridge. I obtained my PhD in 2020 from EPFL under the supervision of Ed Bugnion. My thesis was awarded the 2021 Dennis M. Ritchie Award and the 2021 Honourable Mention for the Eurosys Roger Needham PhD Award. My PhD was supported by an IBM Fellowship and the Microsoft Swiss JRC. I did my undergrad in Electrical and Computer Engineering at NTUA.

I'm always looking for motivated PhD students and interns! If you are an Imperial student don't hesitate to reach out to me if you are interested in working on a project beyond the ones listed in the projects portal.

Teaching

Students

Publications

Achieving Microsecond-Scale Tail Latency Efficiently with Approximate Optimal Scheduling
Rishabh Iyer, Musa Unal, Marios Kogias, George Candea
in SOSP 2023 Paper
When Concurrency Matters: Behaviour-Oriented Concurrency
Luke Cheeseman, Matthew J. Parkinson, Sylvan Clebsch, Marios Kogias, Sophia Drossopoulou, David Chisnall, Tobias Wrigstad, Paul Liétar
in OOPSLA 2023 Paper
HEELS: A Host-Enabled eBPF-Based Load Balancing Scheme
Rui Yang, Marios Kogias
in eBPF'23@SIGCOMM Paper
Creating Trust by Abolishing Hierarchies
Charly Castes, Adrien Ghosn, Neelu S. Kalani, Yuchen Qian, Marios Kogias, Mathias Payer, Edouard Bugnion
in HotOS 2023 Paper
Towards (Really) Safe and Fast Confidential I/O
Hugo Lefeuvre, David Chisnall, Marios Kogias, Pierre Olivie
in HotOS 2023 Paper
When Idling is Ideal: Optimizing Tail-Latency for Heavy-Tailed Datacenter Workloads with Perséphone
Henri Maxime Demoulin, Joshua Fried, Isaac Pedisich, Marios Kogias, Boon Thau Loo, Linh Thi Xuan Phan, Irene Zhang
in SOSP 2021 Paper
Harvesting Idle CPUs Safely and Efficiently in the Cloud
Yawen Wang, Kapil Arya, Marios Kogias, Manohar Vanga, Aditya Bhandari, Neeraja J. Yadwadkar, Siddhartha Sen, Sameh Elnikety, Christos Kozyrakis, and Ricardo Bianchini
in Eurosys 2021 Paper
Enclosure: language-based restriction of untrusted libraries
Adrien Ghosn, Marios Kogias, James Larus, Ed Bugnion
in ASPLOS 2021 Paper
Benchmarking, analysis, and optimization of serverless function snapshots
Dmitrii Ustiugov, Plamen Petrov, Marios Kogias, Ed Bugnion, Boris Grot
in ASPLOS 2021 Distinguished Artifact! Paper
Operating System and Network Co-Design for Latency-Critical Datacenter Applications
PhD Thesis link
Bypassing the Load Balancer without Regrets
Marios Kogias, Rishabh Iyer, Ed Bugnion
in SoCC 2020 Paper
Tail-Tolerance as a Systems Principle not a Metric
Marios Kogias, Ed Bugnion
in APNet2020 Paper
HovercRaft: Achieving Scalability and Fault-tolerance for microsecond-scale Datacenter Services
Marios Kogias, Ed Bugnion
in Eurosys 2020 Best Student Paper Award! Paper
R2P2: Making RPCs first-class datacenter citizens
Marios Kogias, George Prekas, Adrien Ghosn, Jonas Fietz, Ed Bugnion
in Usenix ATC 2019 Paper
Lancet: A self-correcting Latency Measuring Tool
Marios Kogias, Stephen Mallon, Ed Bugnion
in Usenix ATC 2019 Paper
Flow Control for Latency-Critical RPCs
Marios Kogias, Ed Bugnion
in KBNets@SIGCOMM2018 Paper
ZygOS: Achieving Low Tail Latency for Microsecond-scale Networked Tasks
George Prekas*, Marios Kogias*, Ed Bugnion (*equal co-authors)
in SOSP 2017 Paper