5 papers
Evaluating LLM Coding Agents on SZ-Family Lossy Compression Across Architectures
Changqing Li, Shouwei Gao, Kai Zhao +2
Large language model (LLM) coding agents are increasingly applied to code translation and optimization, yet their effectiveness in performance-critical high-performance computing (…
FLYING SERVING: On-the-Fly Parallelism Switching for Large Language Model Serving
Shouwei Gao, Junqi Yin, Feiyi Wang +1
Production LLM serving must simultaneously deliver high throughput, low latency, and sufficient context capacity under non-stationary traffic and mixed request requirements. Data p…
LUMOS: Democratizing SciML Workflows with L0-Regularized Learning for Unified Feature and Parameter Adaptation
Shouwei Gao, Xu Zheng, Dongsheng Luo +2
The rapid growth of scientific machine learning (SciML) has accelerated discovery across diverse domains, yet designing effective SciML models remains a challenging task. In practi…
DOLMA: A Data Object Level Memory Disaggregation Framework for HPC Applications
Haoyu Zheng, Shouwei Gao, Jie Ren +1
Memory disaggregation is promising to scale memory capacity and improves utilization in HPC systems. However, the performance overhead of accessing remote memory poses a significan…
HurriCast: Synthetic Tropical Cyclone Track Generation for Hurricane Forecasting
Shouwei Gao, Meiyan Gao, Yuepeng Li +1
The generation of synthetic tropical cyclone(TC) tracks for risk assessment is a critical application of preparedness for the impacts of climate change and disaster relief, particu…