In September, Meta launched the 2021 Mathematical Modeling and Optimization for Large-Scale Distributed Systems request for proposals (RFP). Today, we’re announcing the winners of this award. This was the first opportunity for academics in this space to submit proposals, and we were pleased to see a positive response.

In this opportunity, we sought proposals that would advance mathematical techniques with a clear, practical application to one area of interest in back-end systems. Areas of interest for this RFP included the following:

• Data center and hardware operations
• Cloud computing and cluster management
• Application tuning and optimization

The RFP attracted 64 proposals from 45 universities and institutions around the world. Thank you to everyone who took the time to submit a proposal, and congratulations to the winners.

Research award recipients

Core techniques and algorithms for scalable geo-distributed data analytics
Liting Hu (Virginia Polytechnic Institute and State University)

Modeling and optimization of data centers
Nicholas Bambos (Stanford University)

Reinforcement learning for online service placement in cloud management
Tian Lan, Vaneet Aggarwal (George Washington University)

Supporting reliable net zero data center operations with network design
Ho-Yin Mak (Oxford University)

System-aware large-scale distributed optimization: Service-oriented designs
Mingyi Hong, Jie Ding, Prashant Khanduri, Zhi-Li Zhang (University of Minnesota Twin Cities)

Systematic testing of high-speed RDMA networks
Danyang Zhuo (Duke University)

Finalists

A distributed agent-based model for optimizing data center operations
Giovanni Iacca, Leonardo Lucio Custode (University of Trento)

Distributed scalable load balancing under data locality
Debankur Mukherjee, He Wang (Georgia Institute of Technology)

Fast explainable many-objective optimisation of software defined networks
David John Walker, Asiya Khan, Matthew Craven, Vasilios Kelefouras (University of Plymouth)

Flexible distributed optimization framework for heterogeneous networks
Ermin Wei (Northwestern University)

Graph-based dynamic programming for large-scale infrastructure networks
Richard Y Zhang, Gavin Zhang (University of Illinois Urbana-Champaign)

Large ML cluster scheduling via multi-agent graph reinforcement learning
Chuan Wu (University of Hong Kong)

Learned flash caching with extended endurance
Yue Cheng, Mingrui Liu (George Mason University)

Mean field game inspired modeling and resource optimization for MEC
Zhu Han, Miao Pan (University of Houston)

Optimal scheduling for parallel and interdependent jobs in data centers
Bin Li, Siva Theja Maguluri (Pennsylvania State University)

Practical traffic engineering for throughput-optimized data centers
Brighten Godfrey (University of Illinois Urbana-Champaign)

Scalable Riemannian optimization for large-scale systems
Shiqian Ma, Lingzhou Xue (University of California, Davis)

The use of neural networks in large-scale hardware failure prediction
Bhiksha Raj, Joseph Konan (Carnegie Mellon University)