Georgia Institute of Technology
Modeling Modern-day Datacenter Design and Performance Using Formal Verification
Abstract
dc:description.abstractModern cloud and datacenter infrastructure is becoming increasingly complex, with diverse workloads, heterogeneous hardware and software choices, and strict performance objectives. Addressing this challenge requires methods that are both rigorous and practical. My work extends a lightweight automated reasoning framework for modeling, optimizing, and explaining workload-specific deployment configurations using formal abstractions and SMT-based verification. Specifically, I extend this framework by understanding the role of LLMs and then incorporating them in a carefully bounded supporting role for tasks such as specification extraction, summarization, and natural-language interaction, while preserving formal reasoning as the source of correctness and guarantees. In addition, I extend formal reasoning beyond design synthesis to quantitative performance verification, showing how the interaction of complex systems sharing resources can be analyzed under worst-case conditions. Together, these contributions demonstrate a path toward more rigorous, explainable, and usable tools for the design and verification of modern cloud and datacenter systems.
Degree
thesis:*- Name thesis:degree_name
- Computer Science, MS
- Department thesis:degree_department
- Computer Science
- Grantor
- Georgia Institute of Technology
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sahu, Pratyush
- Advisor dc:contributor.advisor
-
- Saeed, Ahmed
- Committee members dc:contributor.committeemember
-
- Dainotti, Alberto
- Dhekne, Ashutosh
- ORCID person:identifier.orcid
Rights
- Language dc:language.iso
- English
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/81603