News & Event


Home > News & Event > Seminar

News & Event


Rearchitecting Linux Storage Stack for Remote SSD access


Rearchitecting Linux Storage Stack for Remote SSD access


Feb.10 (Friday) 13:30 


Prof. Jaehyun Hwang




O Speaker: Jaehyun Hwang(Department of Semiconductor Systems Engineering at Sungkyunkwan University)

O Title: Rearchitecting Linux Storage Stack for Remote SSD access

O Date: Feb.10 (Friday) 

O Start Time: 13:30 

O Venue: Zoom(


O Abstract:

There is a widespread belief in the community that it is not possible to achieve high throughput and µs-scale tail latency when using the Linux kernel stack.

Two most frequently cited arguments are (1) Linux has high CPU overheads; and (2) the resource multiplexing principle is so firmly entrenched in Linux that its performance stumbles when
multiple applications compete for host resources.

I will demonstrate that the above belief may be misplaced, at least for remote storage (SSD) access. In particular, I will present a new Linux storage stack architecture that achieves (1) high throughput comparable to NVMe-over-RDMA and (2) µs-scale latency, even when tens of applications compete for host resources while performing operations at throughput close to hardware capacity. Such a performance can be achieved without any modification in applications, network hardware, kernel CPU schedulers and/or kernel network stack.

O Bio:
Jaehyun Hwang is an Assistant Professor in the Department of Semiconductor Systems Engineering at Sungkyunkwan University (SKKU).
Prior to SKKU, he was a Postdoctoral Associate in the Department of Computer Science at Cornell University from 2018 to 2021.
He also worked at Samsung Electronics as a Senior Engineer from 2015 to 2018 and Bell Labs Seoul, Alcatel-Lucent (currently Nokia), as a Member of Technical Staff from 2010 to 2015.
He received his Ph.D. degree in computer science from Korea University (2010). His research interests are generally in system software and operating systems, including storage disaggregation, data center systems, and CPU-efficient storage/network stacks.