Supervisor: Pınar Tözün


Proposals


Current


Previous

Efficient OS-level Context-Switching for Thread Migration Leveraging Heterogeneous Hardware Micro-architectural Analysis of SystemML What is HTAP? Workload Characterization for Big Data Management Analysis of New Generation SSDs Resource management for Data-Intensive Systems Running on Heterogeneous Hardware Unity DOTS Profiling Survey of hash-based authentication on modern hardware Analysis of NVMe SSDs and the IO stack Workload Characterization for Machine Learning Hardware Utilization Estimation with Machine Learning Data Preprocessing Pipelines Disk Access Tracing for Data-Intensive Systems Profiling Infrastructure for DAPHNE Studying Collocation for Machine Learning Managing Experimental Data with DuckDB Flexible Data Placement on SSDs for Database Systems Incremental Model Updates on Tiny Hardware Machine Learning Models for the Edge Benchmarking Edge Devices for Data-Intensive Applications BLOX for Deep Learning Task Scheduling with GPU Collocation Checkpointing during Deep Learning Training GPU Memory Dataset with a focus on Transformer-Based Models Predicting GPU utilization for Deep learning training Alternative IO backends for Database Systems and SSDs Benchmarking Edge Devices for Data-Intensive Applications Data Attribution on Progressive Datasets for Deep Learning Efficient Data Selection Methods for Machine Learning Framework for Systematic Performance Experiments for Machine Learning Going Beyond Memory with GPU-based Data Analytics Resource Management on Tiny Hardware Evaluating the Impact of Collocating Deep Learning Training Tasks on Jetson Orion Nano GPUs