Georgia Institute of Technology
The Application of Modern Black-box Optimization Algorithms to Problems in SiGe HBT Device Design
Abstract
dc:description.abstractThe general goal of this dissertation is to study the application of Bayesian Optimization algorithms to the design of the SiGe HBT. We do this largely through computational means, colloquially known as TCAD. These tools serve to solve the semiconductor equations and offer a means to engineer transistors, diodes, or other semiconducting devices without expensive and time consuming fabrication. Chapter 1 provides a brief introduction and offers context for the work. Chapter 2 treats theoretical and computational modeling of transport in semiconductors, with a focus on model assumptions and limitations. This is critical as, unlike in circuit design, it is of paramount importance for a device engineer to understand the inner workings of the simulator. Chapter 3 provides a theoretical background for mathematical optimization, as well as details on the Bayesian optimizer we apply in this work. Chapter 4 presents investigations of the Ge profile design in the SiGe base. The evaluation efficiency of the optimization approach is validated. The automatic design of temperature stable profiles is then investigated. Chapter 5 investigates automated design of the collector profile. We first present results showing an improved vertical superjunction. We then present results detailing the Pareto front discovery, or the performance bound discovery, of the collector at cryogenic temperatures. Chapter 6 provides conclusions and future work.
Degree
thesis:*- Name thesis:degree_name
- Electrical and Computer Engineering, PhD
- Department thesis:degree_department
- Electrical and Computer Engineering
- Grantor
- Georgia Institute of Technology
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Heimerl, Justin
- Advisor dc:contributor.advisor
-
- Cressler, John D.
- Committee members dc:contributor.committeemember
-
- Yuan, Jiahui
- Doolittle, William Alan
- Sathe, Visvesh
- Niu, Guofu
- ORCID person:identifier.orcid
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/1853/82108