- Found that Linux’s conventional scheduling signals—CPU load and static topology/locality—break down on chiplet CPUs: under shared I/O-chiplet congestion, topologically local placement can run 4.3–4.5× slower than remote placement
- Isolated this “I/O Chiplet Wall” from memory-controller saturation, showing that the slowdown originates from contention along the compute-to-I/O-chiplet path itself
- Built cSwitch, which uses commodity performance counters to infer each thread’s communication path and bandwidth demand, then schedules based on runtime path congestion rather than static topology
Seunghyun An
CV · Google Scholar · GitHub · LinkedIn
Hi, I'm Seunghyun. I am a PhD candidate at the University of Wisconsin–Madison, where I focus on operating systems, chiplet-based server systems, and performance engineering. Currently I am advised by Professor Ming Liu. At KAIST, I worked with Professor Youngjin Kwon on Rust OS and SmartSSDs. At EPFL, I worked with Professor Sanidhya Kashyap on Rust filesystems.
I am seeking Summer 2027 internships in systems software, operating systems, and performance engineering.
Outside research, I am an amateur astronomer interested in visual observing. I also enjoy cooking and spending time outdoors.
seunghyun at cs.wisc.edu