Transforming computing infrastructure.

Cornell's Systems research group advances the fundamental design and implementation of computing infrastructure, from cloud platforms to networking protocols. Researchers develop innovative solutions for distributed systems, fault tolerance, and high-performance computing, shaping how modern digital systems operate and scale.

Faculty exploring systems and networking.

Yifan Wang
Yifan Wang
Ph.D. student, Computer Science
Yifan Wang
Ph.D. student, Computer Science
yw2399@cornell.edu
A photo of Hakim Weatherspoon, a mwan with buzzed black hair and beard and mustache, in a blue shirt and black jacket, in a hallway
Hakim Weatherspoon
Associate Dean for Belonging at Bowers, Professor of Computer Science
Hakim Weatherspoon
Associate Dean for Belonging at Bowers, Professor of Computer Science
hweather@cs.cornell.edu
Zhiru Zhang
Zhiru Zhang
Professor of Electrical and Computer Engineering
Zhiru Zhang
Professor of Electrical and Computer Engineering
zz284@cornell.edu

Core Focus Areas

Pioneering work in virtualization, datacenter applications, and high-performance distributed systemsResearch projects include: 

Virtualization
Advanced nested virtualization techniques enable dynamic migration between cloud providers and efficient consolidation of virtual machines, improving performance through innovative kernel/user space management. (Weatherspoon, van Renesse)

Escher
A novel communication framework for datacenter applications that abstracts distributed system complexities, allowing replicated servers to interact seamlessly without knowledge of their underlying distributed architecture. Alvisi, van Renesse) 

Ziplog 
Datacenter-scale logging service that overcomes traditional scale-latency trade-offs, providing ultra-low latency and seamless reconfiguration for dynamic scaling without performance degradation. (Alvisi, van Renesse)

Derecho (Birman)
High-performance library for cloud computing systems, leveraging RDMA hardware and refined state machine replication to achieve superior speeds in distributed coordination and data replication. (Birman) Find more information about Derecho at http://www.cs.cornell.edu/ken/Derecho-Project

Cascade 
Platform optimizing AI/ML workloads by reducing data copying overhead and enabling efficient resource utilization, particularly for applications requiring low-latency responses. Integrates with major AI frameworks while maintaining high performance. (Birman). Find more information about Cascade at http://www.cs.cornell.edu/ken/Cascade-Project

Leading research in reliable distributed systems, including breakthrough work in state machine replication and chain replication protocols. Current focus includes self-configuring systems and provably correct blockchain protocols. Projects include: 

Core Replication System
Leading innovations in fault-tolerant distributed systems, from Chain Replication (van Renesse, Schneider) used in Microsoft Azure to foundational work on State Machine Replication (Schneider). Current research advances self-configuring replication protocols and provably correct blockchain systems (Morrisett, van Renesse).

Industry Impact
Research contributions shape modern distributed systems, with technologies like the Isis toolkit (Birman) deployed extensively in industry applications. Cornell's theoretical frameworks and practical implementations continue to influence how large-scale systems handle failures and maintain reliability.

Developing next-generation solutions for network programming, cloud security, and high-speed networking infrastructure. Projects span enterprise network management, fault-tolerant applications, and precise network measurements: 

Network Programming
Frenetic develops high-level languages for programming enterprise network switches, enabling modular reasoning about network properties and simplified management of distributed switches. (Foster)

Cloud Network Security
The Nebula Project addresses cloud security challenges through innovations like TCPR, which enables transparent recovery of TCP connections after failures. (Birman, van Renesse, Weatherspoon)

Network Measurement and Reliability
SoNIC provides precise measurements of optical networks through software-defined network interface cards, while parallel work advances high-availability solutions for backbone network routers in collaboration with industry partners. (Birman, van Renesse)