Our paper titled PINCH: An Adversarial Extraction Attack Framework for Deep Learning Models has been published on arXiv.
New paper on adversarial attacks on AI systems
Awarded InnovateUK grant on AI security research commercialization
Peter Garraghan has been awarded an InnovateUK grant for CyberASAP Phase 2. This grant aims to commercialize EDS research into security for AI.
Team member joins Huawei Research, UK
Congratulations to PhD student Borowiec in joining the systems lab Huawei Research, UK as a Research Intern.
Accepted IEEE CLOUD paper on Deep Learning Datacenter Auto-tuning
Our paper titled Trimmer: Cost-Efficient Deep Learning Auto-tuning for Cloud Datacenters has been accepted for publication within the IEEE International Conference on Cloud Computing (CLOUD’22).
Awarded EC funding on next generation memory technology
Peter Garraghan along with his Physics colleague Prof. Manus Hayne have been awarded a European Commission (EC) research grant titled MSI: Ultralow-power, Non-volatile, Random Access Memory Arrays for Data centers and Space Applications (ULTRARAM).
This £362k project aims to design ULTRARAM: the next generation of memory architecture, and evaluate it’s effectiveness within the datacenter and space industries.
PhD awarded in Deep Learning Clusters
Dr. Gingfung Yeung has successfully defended his PhD thesis titled Proactive Interference-aware
Resource Management in Deep Learning Training Clusters, and will be joining Huawei Technologies R&D in Edinburgh as a Research Scientist. Congratulations!
Awarded InnovateUK grant on commercializing secure AI research
Peter Garraghan has been awarded an InnovateUK grant for CyberASAP Phase 1. This grant aims to commercialize EDS’ research into security for AI.
Team member joins Huawei Research, UK
PhD student Damian Borowiec will join the Systems Infrastructure Research (SIR) lab in the Huawei Edinburgh Research Center this Summer as an intern. Congratulations!
New team member: Weijia Li
The EDS Lab welcomes Weijia Li to the group, who will be joining as a new Masters by Research student in digital sustainability.
Accepted IEEE TSC paper on Stragglers
Our paper titled START: Straggler Prediction and Mitigation for Cloud Computing Environments using Encoder LSTM Networks has been accepted for publication within IEEE Transactions on Services Computing.
This work was completed in collaboration with Imperial College London, The University of Melbourne, and Queen Mary University of London.