Harry Xu

发表论文 29 篇 · 总被引 144 次 · h-index 7

代表论文

• Parcae: Proactive, Liveput-Optimized DNN Training on Preemptible Instances (2024 · Symposium on Networked Systems Design and Implementation · 被引 47)
• Prism: Cost-Efficient Multi-LLM Serving via GPU Memory Ballooning (2025 · 被引 26)
• A Tale of Two Paths: Toward a Hybrid Data Plane for Efficient Far-Memory Applications (2024 · USENIX Symposium on Operating Systems Design and Implementation · 被引 23)
• Accelerating Graph Mining Systems with Subgraph Morphing (2023 · European Conference on Computer Systems · 被引 18)
• ConServe: Harvesting GPUs for Low-Latency and High-Throughput Large Language Model Serving (2024 · arXiv.org · 被引 16)
• DRust: Language-Guided Distributed Shared Memory with Fine Granularity, Full Transparency, and Ultra Efficiency (2024 · USENIX Symposium on Operating Systems Design and Implementation · 被引 14)