Scalable Dynamic Deployment of Data-Flow Applications in Heterogeneous Environments through AI-Driven Optimization
The increasing complexity of scientific workflows and the growing diversity of computational platforms (HPC, cloud, edge) have made it challenging to efficiently deploy and scale data-flow applications. This PhD research aims to leverage AI-driven optimization techniques to enable dynamic and adaptive deployment of data-flow applications across heterogeneous environments.
Primary Supervisor: Dr Rosa Filgueira
Overview of research area
The research work will focus on developing novel techniques for the efficient and scalable deployment of scientific workflows. By integrating AI-driven heuristics, this research will address the current challenges of dynamic deployment, resource optimization, and autoscaling, which are essential for handling large-scale scientific applications in diverse computing environments. Building on frameworks like dispel4py [1,2] and Laminar[2,3], this research will develop advanced models and algorithms to enhance adaptability and efficiency.
Potential research question(s)
- How can AI-driven heuristics improve the efficiency and adaptability of data-flow applications in heterogeneous environments?
- What machine learning techniques [5] can predict resource demands for real-time autoscaling?
- How can execution portability be maintained across diverse platforms while ensuring optimal performance?
Student Requirements
A UK 2:1 honours degree, or its international equivalent, in a relevant subject such as computer science and informatics, physics, mathematics, engineering, biology, chemistry and geosciences.
You must be a competent programmer in at least one of C, C++, Python, Fortran, or Java and should be familiar with mathematical concepts such as algebra, linear algebra and probability and statistics.
English Language requirements as set by University of Edinburgh.
Student Recommended/Desirable Skills and Experience
- Experience with AI/ML techniques and deployment strategies.
- Knowledge of HPC or cloud computing environments.
- Understanding of data-flow programming models and workflows.
How to apply
Applications should be made via the University application form, available via the degree finder. Please note the proposed supervisor and project title from this page and include this in your application. You may also find this page is an uneful starting point for a research proposal and we would strongly recommend discussing this further with the potential supervisor.
References
[1] Optimization towards Efficiency and Stateful of dispel4py
[3] Laminar: A New Serverless Stream-based Framework with Semantic Code Search and Code Completion
[5] Auto-Scaling Techniques in Cloud Computing: Issues and Research Directions