About Me


Short Bio:

I completed my Ph.D. in Electrical and Computer Engineering at WPI, where I was advised by Prof. Berk Sunar and Prof. Thomas Eisenbarth. I focused on the security of cryptographic implementations and micro-architectural systems including Intel and ARM processors. Due to the increasing variety of the micro-architectural attacks and huge amount of data, I worked on several Deep Learning algorithms to improve the existing attacks. In addition, I conducted experiments on detecting state-of-the-art micro-architectural attacks such as Spectre, Meltdown, Zombieload by using Deep Learning algorithms. I was the recipient of Global Research Award from WPI in 2017. I focused on cache attacks in ARM devices at Fraunhofer AISEC, Munich. I also had an internship at VMware Security Group, where I analyzed the possibility of side-channel attacks in the cloud environment.


  • FastSpec: Scalable Generation and Detection of Spectre Gadgets Using Neural Embeddings
    MC Tol, K Yurtseven, B Gulmezoglu, B Sunar   
    Arxiv eprint (PDF)
  • FortuneTeller: Predicting Microarchitectural Attacks via Unsupervised Deep Learning 
    B Gulmezoglu, A Moghimi, T Eisenbarth, B Sunar   
    Arxiv eprint (PDF) (Bib)
  • SPOILER: Speculative Load Hazards Boost Rowhammer and Cache Attacks
    S Islam, A Moghimi, I Bruhns, M Krebbel, B Gulmezoglu, T Eisenbarth, B Sunar   
    USENIX Security 2019 (PDF) (Bib)
  • PerfWeb: How to violate web privacy with hardware performance events
    B Gulmezoglu, A Zankl, T Eisenbarth, B Sunar
    ESORICS 2017 (PDF) (Bib)
  • Cache attacks enable bulk key recovery on the cloud
    MS Inci, B Gulmezoglu, G Irazoqui, T Eisenbarth, B Sunar
    CHES 2016 (PDF) (Bib)


  • VMware Product Security Team May 2015 – August 2015
    I worked on reverse-engineering the TLB structures in several micro-architectures. Then, I created some experiments for TLB-based side channel attacks in VMware products.
  • ASELSAN Communication and Information Technologies Inc. July 2010 – August 2010
    I worked on Friend and Foe target detection by satellites.


  • Fraunfoher AISEC October 2017 – December 2017
    I focused on micro-architectural attacks on mobile platforms. I developed an application which can monitor cache activities in high performance cores. The collected information is used to classify the activities of the user with the help Convolutional Neural Networks.