Jeongseob Ahn
where else to find them
no profile links on file
what they talk about
where they keep going back
no series more than once — all 4 in the filters
on a stage with
all 13 in the filters
who talks about the same things
nobody close enough yet
| title | conference | date |
|---|---|---|
| Revisiting Pipeline Parallelism for LLM Serving | 20th USENIX Symposium on Operating Systems Design and Implementation (OSDI '26) | 2026-07-13 |
Revisits pipeline parallelism with an implementation on SGLang that runs on four NVIDIA A100 40GB GPUs and evaluates Qwen2.5 32B and 14B. Introduces greedy and… | ||
| EnvPipe: Performance-preserving DNN Training Framework for Saving Energy | USENIX ATC '23 | 2023-07-10 |
EnvPipe is a PyTorch library that profiles each pipeline stage, reschedules pipeline units to expose usable pipeline bubbles, and selectively lowers GPU SM… | ||
| Memory Harvesting in Multi-GPU Systems with Hierarchical Unified Virtual Memory | USENIX ATC '22 | 2022-07-11 |
Implements a hierarchical unified virtual memory (HUVM) that spills GPU pages to neighbor GPUs over NVLink instead of host PCIe, and ships a centralized memory… | ||
| Zico: Efficient GPU Memory Sharing for Concurrent DNN Training | USENIX ATC '21 | 2021-07-14 |
Zico builds a shared GPU memory pool plus a memory manager that uses CUDA events and an in‑flight kernel counter to safely reclaim and limit allocations, and a… | ||
| Exploring the Design Space of Page Management for Multi-Tiered Memory Systems | USENIX ATC '21 | 2021-07-14 |
Linux 5.3 is modified and evaluated on a two-socket system combining DRAM and Intel DCPMM to expose page placement choices. The design implements… | ||
| nothing matches | ||