Efficient AI Systems
Faster and fairer large language model inference across heterogeneous edge resources.
School of Computer Science · University of Sydney
Distributed intelligence for real‑world systems.
We develop efficient and reliable learning systems at the intersection of distributed computing, optimization, and artificial intelligence.
About DUAL
From mathematical foundations to deployable systems.
The Distributed compUting, optimizAtion, and Learning (DUAL) Group connects rigorous theory with intelligent systems that can operate at scale.
We study resource-constrained environments where communication, compute, fairness, and reliability matter just as much as model accuracy.
Research themes
Building dependable intelligence, together.
Federated & Edge Learning
Collaborative learning that respects resource, privacy, and communication constraints.
Distributed Optimization
Scalable algorithms with robust convergence and practical systems performance.
Semantic Communication
Reliable meaning-aware communication under uncertainty and changing wireless conditions.
Distributed Speculative Decoding for Edge LLM Inference
Coordinated decoding systems that accelerate language model inference across heterogeneous edge infrastructure.
Featured projects
Research in motion.
Robust communication WaSeCom Distributionally robust wireless semantic communication with large AI models. 
Efficient inference GOODSPEED Speculative decoding for fast and fair LLM inference at the edge. Selected publications
Recent work from the group.
People & collaboration
Curious minds are welcome.
DUAL brings together researchers and students across machine learning, distributed systems, optimization, and edge intelligence.
