Youjip Won
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Youjip Won has a total of 21 co-authors. Here are the top co-authors by number of shared publications:
- Juwon Kim
KAIST, Daejeon, Republic of Korea
3 shared publications - Erez Zadok
Stony Brook University, Stony Brook, NY, USA
2 shared publications - Geoff Kuenning
Harvey Mudd College, Claremont, CA, USA
2 shared publications - Hwanseok Yoo
KAIST, Daejeon, Republic of Korea
2 shared publications - Seung Won Yoo
KAIST, Daejeon, Republic of Korea
2 shared publications - Daejun Park
Samsung Electronics
1 shared publication
Latest Publications
- Open Access
- 28 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 321-336
GAZE: Rearchitecting Log-Structured Filesystem for Garbage Collection on Zoned Namespace Flash Storage
- 28 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 321-336
Open AccessGAZE: Rearchitecting Log-Structured Filesystem for Garbage Collection on Zoned Namespace Flash Storage
- Open Access
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 92-98
Manycore Scalable Logging for Log-Structured Filesystem
- 27 September 2026
- Published by Association for Computing Machinery (ACM)
- p. 92-98
Open AccessManycore Scalable Logging for Log-Structured Filesystem
- Open Access
- 26 February 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 22 (3) , 1-39
Advanced I/O Stack for Efficient Zone-based Mobile Flash Storage
We propose an I/O stack for ZNS based flash storage in mobile environment, ZMS . The zone interface is known to save the flash storage from two fundamental issues which modern flash storage suffers from: logical-to-physical mapping table size and garbage collection overhead. Through extensive study, we find that realizing the zone interface in mobile environment is more challenging due to the unique characteristics of mobile environment: the lack of on-device memory in mobile flash storage and the frequent fsync() calls in mobile applications. Aligned with this, we identify the root causes that need to be addressed in realizing the zone interface in mobile I/O stack: write buffer thrashing and tiny synchronous file update . We develop a file system, block I/O layer, and device firmware techniques to address the above mentioned two issues. The three key techniques in ZMS are (i) IOTailor, (ii) budget-based in-place update, and (iii) multi-granularity logical-to-physical mapping. We also implement copy offloading and propose a new overprovisioned space configuration for the F2FS file system to address the host-side garbage collection overhead. Evaluation on a real production platform shows that ZMS improves the write amplification by 2.9–6.4 × and the random write performance by 5.0–13.6 ×. With the three techniques, ZMS shows a significant performance improvement in writing the data blocks to the multiple zones concurrently, in executing SQLite transactions, in launching the applications, and in playing video files.
- 26 February 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 22 (3) , 1-39
Open AccessAdvanced I/O Stack for Efficient Zone-based Mobile Flash Storage
We propose an I/O stack for ZNS based flash storage in mobile environment, ZMS . The zone interface is known to save the flash storage from two fundamental issues which modern flash storage suffers from: logical-to-physical mapping table size and garbage collection overhead. Through extensive study, we find that realizing the zone interface in mobile environment is more challenging due to the unique characteristics of mobile environment: the lack of on-device memory in mobile flash storage and the frequent fsync() calls in mobile applications. Aligned with this, we identify the root causes that need to be addressed in realizing the zone interface in mobile I/O stack: write buffer thrashing and tiny synchronous file update . We develop a file system, block I/O layer, and device firmware techniques to address the above mentioned two issues. The three key techniques in ZMS are (i) IOTailor, (ii) budget-based in-place update, and (iii) multi-granularity logical-to-physical mapping. We also implement copy offloading and propose a new overprovisioned space configuration for the F2FS file system to address the host-side garbage collection overhead. Evaluation on a real production platform shows that ZMS improves the write amplification by 2.9–6.4 × and the random write performance by 5.0–13.6 ×. With the three techniques, ZMS shows a significant performance improvement in writing the data blocks to the multiple zones concurrently, in executing SQLite transactions, in launching the applications, and in playing video files.
- Open Access
- 6 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Embedded Computing Systems
- Vol. 25 (1) , 1-27
SeeSSD: Computational Storage for Energy-Efficient Real-Time Object Detection
In this work, we present our intelligent SSD, SeeSSD , an energy-efficient computational SSD for a real-time object detection system. SeeSSD embeds an FPGA-based CNN processing engine and the firmware that performs the convolutional operation on the target image. SeeSSD processes the image data at the storage before sending it to the host. This reduces the amount of data transferred to the host and lowers the data movement overhead, thus reducing transfer time and saving power. By using our SeeSSD system and YOLO_Embed, an object detection neural network model, we are able to outperform the fastest YOLO model for an embedded controller, YOLO-Lite, in terms of performance, accuracy, and energy efficiency. YOLO (You Only Look Once) models are a series of one-stage object detection neural models that have become very popular due to their fast speed and high accuracy. The contribution of this work includes designing and implementing our SeeSSD system with a lightweight object detection model, YOLO_Embed, for reducing the data movement overhead, performing real-time inference, and lowering the overall power consumption. We implemented the entire software stack associated with the SeeSSD system; on-device CNN acceleration engine implemented on FPGA, object identification interface for SeeSSD using YOLO_Embed, and embedded software layer in SeeSSD for on-device convolutional processing. We calculated our YOLO_Embed model’s accuracy on object detection dataset benchmarks such as PASCAL VOC 2012, which came out to be 38.1% mAP (mean Accuracy Precision). Our system was able to perform inference in 0.21 seconds while reducing the power consumption by approximately 1.2× and 1.4× for CPU-Only and CPU+GPU systems, respectively. We were also able to reduce the data movement overhead by 24× for a single target image.
- 6 January 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Embedded Computing Systems
- Vol. 25 (1) , 1-27
Open AccessSeeSSD: Computational Storage for Energy-Efficient Real-Time Object Detection
In this work, we present our intelligent SSD, SeeSSD , an energy-efficient computational SSD for a real-time object detection system. SeeSSD embeds an FPGA-based CNN processing engine and the firmware that performs the convolutional operation on the target image. SeeSSD processes the image data at the storage before sending it to the host. This reduces the amount of data transferred to the host and lowers the data movement overhead, thus reducing transfer time and saving power. By using our SeeSSD system and YOLO_Embed, an object detection neural network model, we are able to outperform the fastest YOLO model for an embedded controller, YOLO-Lite, in terms of performance, accuracy, and energy efficiency. YOLO (You Only Look Once) models are a series of one-stage object detection neural models that have become very popular due to their fast speed and high accuracy. The contribution of this work includes designing and implementing our SeeSSD system with a lightweight object detection model, YOLO_Embed, for reducing the data movement overhead, performing real-time inference, and lowering the overall power consumption. We implemented the entire software stack associated with the SeeSSD system; on-device CNN acceleration engine implemented on FPGA, object identification interface for SeeSSD using YOLO_Embed, and embedded software layer in SeeSSD for on-device convolutional processing. We calculated our YOLO_Embed model’s accuracy on object detection dataset benchmarks such as PASCAL VOC 2012, which came out to be 38.1% mAP (mean Accuracy Precision). Our system was able to perform inference in 0.21 seconds while reducing the power consumption by approximately 1.2× and 1.4× for CPU-Only and CPU+GPU systems, respectively. We were also able to reduce the data movement overhead by 24× for a single target image.
- Open Access
- 22 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (3) , 1-3
The Past, Present, and Future of Storage Technologies (part 2 of 2)
- 22 July 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (3) , 1-3
Open AccessThe Past, Present, and Future of Storage Technologies (part 2 of 2)
- Open Access
- 31 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1374-1382
lwFSCK: Light-weight Filesystem Check
- 31 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 1374-1382
Open AccesslwFSCK: Light-weight Filesystem Check
- Open Access
- 22 February 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (2) , 1-2
Introduction to the Special Section on USENIX FAST 2024
- 22 February 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (2) , 1-2
Open AccessIntroduction to the Special Section on USENIX FAST 2024
- Open Access
- 15 January 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (1) , 1-3
The Past, Present, and Future of Storage Technologies (Part 1 of 2)
- 15 January 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Storage
- Vol. 21 (1) , 1-3
Open AccessThe Past, Present, and Future of Storage Technologies (Part 1 of 2)