PI: Nuno Preguiça

The mission of the NOVA LINCS Computer Systems group is to find ways to make the systems that surround us in our daily lives more reliable, trustworthy, dependable, and better performing. This goal is increasingly important, as our societies are increasingly dependent on services that are offered by these computer systems, from e-commerce and social networks to business applications and support for scientific activities. Furthermore, the complexity of these systems is constantly increasing, in particular through the use of cloud services that are supported by an infrastructure running in geographically dispersed data centers, with tens to hundreds of thousands of servers, each with tens of processors. This complexity makes the goal of creating systems that run continuously and correctly ever more challenging. Fulfilling this goal involves techniques that span a broad set of areas, both covered by the group and spanning other groups from NOVA LINCS, e.g., distributed systems and networking, operating systems, parallel and transactional systems, security, programming languages, etc.

Over the last few years, researchers from this group have worked on several problems in the broad area of computer systems, with a particular focus on distributed and parallel systems. Examples include finding principled approaches to building systems that support weak consistency models, namely through the use of data types called CRDTs, which ensure the convergence of all the replicas irrespectively of the order by which they apply updates; devising ways to build geo-replicated systems that make use of efficient weakly consistent algorithms whenever possible but resort to strong consistency models whenever necessary; or developing new methods for programmers to exploit the parallelism in modern computer systems (both concurrent programs running on multi-cores and big data computations running inside data centers), while offering a programming model that is intuitive and simple to use.

The group has a relevant presence in the international community, with results published at elite venues, such as OSDI, EuroSys, NSDI, PODC and ICDCS, and team members frequently serving in the PC and organization of top venues. Group members received prestigious awards, including an ERC starting grant, a Google Research Award, a Fraunhofer Portugal Award, and several best/distinguished paper awards. The group members participate in multiple funded projects, including EU research projects and networks, national research projects and collaborations with companies.