1

Zorse: Optimizing LLM Training Efficiency on Heterogeneous GPU Clusters

Conference on Machine Learning and Systems (MLSys), 2026

Sampling-based Predictive Database Buffer Management

Proceedings of Very Large Data Bases (VLDB) 2025

Cephalo: Harnessing Heterogeneous GPU Clusters for Training Transformer Models

International Conference on Supercomputing (ICS) 2025

Rethinking Closed-Loop Training for Autonomous Driving

European Conference on Computer Vision (ECCV), 2022

Hydrozoa: Dynamic Hybrid-Parallel DNN Training on Serverless Containers

Conference on Machine Learning and Systems (MLSys), 2022

RadarNet: Exploiting Radar for Robust Perception of Dynamic Objects

European Conference on Computer Vision (ECCV), 2020