Hosein Yavarzadeh

I’m a fifth-year PhD candidate in the CSE department at the University of California, San Diego (UCSD).

My research interests lie in computer security and computer architecture, with a particular emphasis on confidential computing, microarchitectural side-channel attacks, and high-performance architectural defenses such as control-flow isolation.

I am advised by Professor Dean Tullsen and Professor Deian Stefan. I actively collaborate with Professor Kazem Taram, Dr. Daniel Moghimi (Google), and collaborators from multiple universities and industry labs.

news

Sep 20, 2025	I’m back at Google for the next six months as a Student Researcher, working on confidential computing security and privacy.
Jun 05, 2025	I’m joining Qualcomm’s CPU Security team as a Research Intern for the next three months, working on security for next-generation Qualcomm CPUs.
Nov 19, 2024	I’m joining Google as a Student Researcher (November–April), working with Daniel Moghimi on confidential computing security and privacy.
Aug 14, 2024	Our paper Indirector has been accepted to USENIX Security 2024 and received the Distinguished Paper Award!
Jun 10, 2024	I’m joining Qualcomm’s CPU Performance Modeling team as a Research Intern, researching novel branch prediction designs to improve the performance of next-generation Qualcomm Snapdragon X Elite CPUs.

selected publications

USENIX Sec

Indirector: High-Precision Branch Target Injection Attacks Exploiting the Indirect Branch Predictor

Hosein Yavarzadeh, Luyi Li, and Dean Tullsen

In 33rd USENIX Security Symposium (USENIX Security 24), Aug 2024

*Joint first authorship

USENIX Security Distinguished Paper Bib HTML

Distinguished Paper Award Winner

@inproceedings{li2024indirector,
  title = {Indirector: {High-Precision} Branch Target Injection Attacks Exploiting the Indirect Branch Predictor},
  author = {Yavarzadeh, Hosein and Li, Luyi and Tullsen, Dean},
  booktitle = {33rd USENIX Security Symposium (USENIX Security 24)},
  year = {2024},
  month = aug,
  address = {Philadelphia, PA},
  pages = {2137--2154},
  publisher = {USENIX Association},
  url = {https://www.usenix.org/conference/usenixsecurity24/presentation/li-luyi},
  note = {*Joint first authorship}
}

ASPLOS

Pathfinder: High-Resolution Control-Flow Attacks Exploiting the Conditional Branch Predictor

Hosein Yavarzadeh, Archit Agarwal, Max Christman, Christina Garman, Daniel Genkin, Andrew Kwong, Daniel Moghimi, Deian Stefan, Kazem Taram, and Dean Tullsen

In International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), Apr 2024

Reverse engineering and exploitation of modern conditional branch predictors

Bib HTML Video

@inproceedings{yavarzadeh2024pathfinder,
  title = {Pathfinder: High-Resolution Control-Flow Attacks Exploiting the Conditional Branch Predictor},
  author = {Yavarzadeh, Hosein and Agarwal, Archit and Christman, Max and Garman, Christina and Genkin, Daniel and Kwong, Andrew and Moghimi, Daniel and Stefan, Deian and Taram, Kazem and Tullsen, Dean},
  booktitle = {International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS)},
  year = {2024},
  month = apr,
  url = {https://mktrm.github.io/files/pathfinder-asplos24.pdf},
  artifact = {https://github.com/hoseinyavarzadeh/Pathfinder_Attack},
  note = {Reverse engineering and exploitation of modern conditional branch predictors}
}

IEEE S&P

Half&Half: Demystifying Intel’s Directional Branch Predictors for Fast, Secure Partitioned Execution

Hosein Yavarzadeh, Mohammadkazem Taram, Shravan Narayan, Deian Stefan, and Dean Tullsen

In 2023 IEEE Symposium on Security and Privacy (S&P), May 2023

High-performance control-flow isolation against branch-prediction side channels

Bib HTML

@inproceedings{yavarzadeh2023halfnhalf,
  title = {Half\&Half: Demystifying Intel’s Directional Branch Predictors for Fast, Secure Partitioned Execution},
  author = {Yavarzadeh, Hosein and Taram, Mohammadkazem and Narayan, Shravan and Stefan, Deian and Tullsen, Dean},
  booktitle = {2023 IEEE Symposium on Security and Privacy (S\&P)},
  year = {2023},
  month = may,
  pages = {1220--1237},
  publisher = {IEEE},
  url = {https://mktrm.github.io/files/halfandhalf_sp23.pdf},
  note = {High-performance control-flow isolation against branch-prediction side channels}
}