Dr Georgios Karakonstantis
The primary focus of Karakonstantis’ research lies on the development of energy-efficient and dependable circuits and systems for portable embedded devices and high-end servers targeting various multimedia, communication, data-mining, health-monitoring, as well as high-performance applications within Edge and Cloud environments.
The group’s work spans three main topics:
- Characterization of the intrinsic variability in memories and design of low cost SW/HW schemes for enabling operation at extended operating points, while addressing potential hardware induced errors.
- Development of error-resilient, approximate and low-power general-purpose and application-specific processors by exploiting application characteristics.
- Implementation of data-processing accelerators and scheduling algorithms for utilizing FPGAs and GPUs in Cloud data-centres and Edge environments.
Open to PhD applications in the field of
- Power and reliability characterisation and machine learning based modelling of processors and memories
- Energy-efficient software, hardware design schemes for dependable/secure systems
- Accelerators (FPGAs, co-processors) in Cloud and Edge environments
- Cross-layer approximate computing based techniques
- Circuits and systems for multimedia, communications and machine learning applications
- Open PhD projects:
- Energy-Quality Scalable Processors based on Data Mining
- A Smart Data Analysis System for Food Safety based on Pattern Recognition
- Failure Prediction based on Data Mining for Energy-Efficient & Dependable Cloud/Edge Ecosystems
- In-Time and In-Space Sharing of FPGA Resources in Cloud/Edge Ecosystems
- Intelligent Memory Access Pattern Scheduling for Power and Performance Optimization
PUBLIC OUTREACH AND KEY ACHIEVEMENTS
PhD title: Energy-Efficient and Variation-Aware Processors by Exploiting Application- Resilience
Name: Ioannis Tsiokanos
Years of Study: 2017 - 2020
PhD title: Design Methods for Exceeding the Energy Scaling Boundaries in Memories
Name: Konstantinos Tovletoglou
Years of Study: 2016 - 2019
PhD title: Exploiting Reconfigurable Architectures in Heterogeneous Edge and Cloud Data Center Environments
Name: Umar Ibrahim Minhas
Years of Study: 2015 - 2020
The Graduate School
Our dedicated hub for postgraduate students, providing state-of-the-art facilities for study, and high-quality training and development programmes within a diverse interdisciplinary community.The gateway to your postgraduate experience