Tyler Estro

Tyler Estro
 Ph.D. Candidate
 File Systems and Storage Lab (FSL)

 Advisor: Professor Erez Zadok
 E: testro [at] cs [dot] stonybrook [dot] edu

 Google Scholar
 CV

 

I'm a Ph.D. student in the Department of Computer Science at Stony Brook University. I'm a member of the File Systems and Storage Lab (FSL) and affiliated with the Institute for AI-Driven Discovery and Innovation. My current research focuses on tiered-memory systems that leverage Compute Express Link (CXL) technology. I’ve published across a diverse range of topics, including the efficient exploration and optimization of multi-tier caching systems, performance modeling, visualization, artificial intelligence, and even applications of high-performance computing and AI in life sciences.

If you see me around, ask me about my research, hobbies, or just say hello and blankly stare. I'm always looking for new opportunities and collaborations. :)

Education

  • Ph.D. in Computer Science - SUNY Stony Brook University, NY. (2018-Current)
  • B.S. in Software Technology - SUNY Farmingdale, NY.
  • A.S. in Business Administration - SUNY Suffolk, NY.

Graduate Coursework

Work Experience

  • Research Assistant: File Systems and Storage Lab, Stony Brook University, NY. (May 2018-Current)
  • High Performance Computing Assistant: Institute for Advanced Computational Science, Stony Brook University, NY. (Aug 2017-May 2018)

Teaching Experience

Publications

Posters

Projects

  • CXL-based live virtual machine migration in QEMU/KVM
    • Designed a CXL-based live migration method that transfers guest memory in a single pass, reducing migration time, blackout duration, and total data transferred compared to network-based techniques
    • Implemented transparent guest VM memory tiering in QEMU/KVM
    • Developed entirely in QEMU/KVM without kernel modifications; evaluated on CXL 2.0 hardware
  • FOUNT: Scaffolded Hands-On Learning for a Data-Centric Future
    • Developed numerous Jupyter-based "courselets", self-contained notebooks for teaching data-systems concepts, deployed on Chameleon Cloud
    • Integrated courselets into a graduate-level course and refined them based on student feedback, producing open educational resource design guidelines
    • Presented a demonstration of courselets during the educational symposium at the Chameleon User Meeting 2023
  • Open CAS tiered-storage workload tracing tool
    • Enhanced Open Cache Acceleration Software (Open CAS) and Linux kernel device drivers to capture low-level traces of multi-tiered caching operations
    • Instrumented kernel event tracing to capture request response times, service times, and inter-arrival times for use in queuing models
  • Multi-tier cache simulator
    • Extended PyMimircache to simulate multi-tier caches
    • Implemented evaluation engine that combines per-device performance characteristics and price data to generate performance and dollar-cost metrics
    • Conducted large-scale simulations using hundreds of real-world workload traces, exploring millions of cache configurations across varied devices and caching parameters

Awards

  • Best Paper Award, 27th IEEE International Symposium on the Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS '19), October 2019, Rennes, France
  • First Place in the "Battle of the Brains" Programming Competition, SUNY Suffolk, NY. (2012)
  • Computer Science, Engineering, and Mathematics Scholarship (CSEMS), SUNY Suffolk, NY. (2012)