Ovidiu-Cristian MARCU
![]() |
|
||||
Department | PCOG | ||||
Postal Address |
Université du Luxembourg Maison du Nombre 6, Avenue de la Fonte L-4364 Esch-sur-Alzette |
||||
Campus Office | MNO, E02 0225-040 | ||||
![]() |
|||||
I studied Computer Science at the Polytechnic University of Bucharest (2004-2009), and then have worked as a Software Engineer for software corporations (e.g., IBM) and startups. In 2015 I had joined the KerData team at Inria Rennes and had worked on my thesis “KerA: A Unified Ingestion and Storage System for Scalable Big Data Processing”, which I defended on 18 December 2018. KerA is a novel system for real-time stream storage and processing that obtains latency and throughput performance improvements, up to an order of magnitude better than state of the art (e.g., Apache Kafka), while leveraging commodity hardware. KerA (see https://kerdata.gitlabpages.inria.fr/Kerdata-Codes/kera-website/) is developed on top of the RAMCloud key-value store and has API integrations with Apache Flink. I have joined the University of Luxembourg on 15 February 2021 as a Postdoctoral Researcher. My main research interests, related to HPC and Big Data convergence, are exploring new system techniques for autonomous (self-tunable), and high-performance, storage and processing systems while leveraging new hardware like persistent memory (e.g., Intel Optane), fast networks (e.g., Infiniband), etc., all while at the microsecond time. We call these Granular Storage Systems, see https://granular-storage.github.io/.
Last updated on: Wednesday, 17 March 2021

2021

; ; ;
in 2021 IEEE International Conference on Cluster Computing (CLUSTER) (2021)