10 papers
From LLM Inference to Agentic Workloads: Characterization and Implications for Serving Systems
Chaokun Chang, Yukun Zhou, Kaihua Fu +10
Agentic applications are shifting AI serving from isolated model inference to long-running workloads in which LLMs coordinate tools, environments, and persistent state. However, th…
ServerlessT2I: Efficient Text-to-Image Workflow Serving on a Serverless Platform
Xiaoxiao Jiang, Suyi Li, Sheng Yao +5
ServerlessT2I breaks down text-to-image generation pipelines into separate model functions that can be independently scheduled on a serverless platform, allowing per-model scaling…
Gleam: Adaptive Network-Efficient CUDA API Remoting for Cross-Device GPU Sharing over LANs
Zhihao Xu, Hao Zhong, Zeting Zhou +9
This paper aims to enable computation- and communication-efficient GPU sharing across devices within local area networks (LANs), facilitating ubiquitous AI inference on heterogeneo…
DisagFusion: Asynchronous Pipeline Parallelism and Elastic Scheduling for Disaggregated Diffusion Serving
Hantian Zha, Teng Ma, Yang Yong +7
Diffusion-based generation is increasingly powering production content pipelines; however, deploying these models at scale remains a significant challenge. Model weights frequently…
Benchmarking Confidential GPU Inference on NVIDIA H100 under Intel TDX
Wei Wang, Abdul Hyee Waqas, Burns Smith
Confidential computing is becoming a practical deployment requirement for AI inference workloads that process sensitive inputs or protect proprietary model assets. However, the per…
Polars inside Intel SGX2 Enclaves: An Empirical Study of Confidential Analytical Query Processing
Wei Wang, Burns Smith, Kenny Leftin
Trusted Execution Environments (TEEs) have renewed interest in confidential analytics, but most prior evaluations focus on SQL database engines or earlier SGX generations. This pap…