OBJECTIVE

I’m a Computer Science Ph.D. student at Virginia Tech. I am broadly interested in computer systems, especially "How to design a scalable, high performance storage systems".
My research interests include but are not limited to, SSD(solid-state device), file system, CXL(compute express link), operating system, and flash-centric systems for AI applications(NDP).

Education Ph.D. in Computer Science Virginia Tech Aug. 2023 -

Master’s Degree in Computer Science Dankook University Graduated in Aug. 2023

Visiting Scholar Syracuse University July – Dec. 2022

Bachelor of Computer Science Dankook University Graduated in Feb. 2022

Patent

Jongmoo Choi, Samuel Woo, and Inho Song Korea: filed
Method for analyzing vehicle forensic and computing device for execution the same 10-2022-0139234

research project

Building a Big Data System with Next-generation SSDfunded by IITP (2022 – 2030)

• This is the StarLab project in IITP. There are numerous studies about next-generation SSD focusing on the bottleneck of the existing storage system. I am expecting to provide higher isolation support and stable tail latency with HW-SW co-design.

• Expected impact:

  1. Achieve optimal throughput with lower garbage collection overhead and lower write amplification factor. (Motivated by F2FS[1] and ZNS[2])

  2. Supports high isolation among tenants with scalable design. (Inspired by MAX[3] and Tectonic[4])

Design Tradeoff in ZNS SSD funded by SK Hynix(2020-2022)

• This is research for ZNS SSD internals. The main goal of this work is to show the design tradeoff in ZNS SSD with a reliable ZNS emulator.

• Impact:

  1. This ZNS emulator shows under 6-17% error rate on average which is accurate performance compared to the real device.Based on this software, 3 main lessons are uncovered which motivates the existing ZNS ecosystem.

  2. This work shows the tradeoff between different designs in ZNS. However, it also uncovers that existing ZNS ecosystems are unaware of this tradeoff in ZNS.

Semantic-aware Vehicle Forensic funded by SPO(2020-2021) and IITP(2022)

• According to Google, modern vehicles are now generating about 1GB of data every second. This emerging field is not only in the spotlight of the forensic investigation field but also for the storage systems. After recognizing significant results, the project grows more important. This work motivated development of a forensic big data system that is useful to SPO and national police organizations.

• Impact:

  1. Based on a holistic approach to retrieving the artifacts from the vehicle, we found critical artifacts that strongly imply the driver’s or accompanies behaviors.

  2. This work shows that even if the driver deletes data in the vehicle, data could still be retrieved by physical dump with metadata analysis.

SKILLS

Programming	C/C++, Python, and Java
Systems	Linux Kernel, RocksDB, and F2FS
Open Source Contribution	FEMU (https://github.com/vtess/FEMU)

Academic Awards and Achievements

• Academic Excellence Scholarship Fall Semester 2022

• Academic Excellence Scholarship Spring Semester 2022

• Graduation Excellence Award Dankook Univ. 2022

• Dean’s List Spring Semester 2020

• Academic Excellence Scholarship Spring Semester 2020

• Dean’s List Fall Semester 2019

• Academic Excellence Scholarship Fall Semester 2016

• Admission Scholarship Spring Semester 2015

Certification

• Teacher’s Certificate Ministry of Education,
  The Secondary School Teacher(Grade II) of Information & Computer Republic of Korea, 2022

Miscellaneous Activities

• International Joint Workshop for High-Potential Individuals Global Training Program IITP 2022

• CPU cache simulator [Git] 2021

• Dankook university data analysis AI contest (Rank: 13/400) DACON 2021

• Camino de Santiago Santiago Compostella, Spain, 2018

• Military service in Republic of Korea Army Republic of Korea, Jan. 2017 – Oct. 2018

  • Special Warrior Certification

• Editor, Consulate General of the Republic of Korea in Jeddah Jeddah, Saudi Arabia, Sep. – Dec. 2016