7 papers
DeltaBox: Scaling Stateful AI Agents with Millisecond-Level Sandbox Checkpoint/Rollback
Yunpeng Dong, Jingkai He, Shiqi Liu +7
LLM-powered AI agents require high-frequency state exploration (e.g., test-time tree search and reinforcement learning), relying on rapid checkpoint and rollback (C/R) of the compl…
ParaCell: Paravirtualized Secure Containers with Lightweight Intra-Container Isolation and Intent-Driven Memory Management
Yiyang Wu, Xunjie Wang, Jinyu Gu +1
Secure containers isolate each container with its own kernel, mitigating shared-kernel attacks prevalent in traditional container systems. However, existing designs still face a fu…
A Case for Agentic Tuning: From Documentation to Action in PostgreSQL
Hongyu Lin, Mingyu Li, Weichen Zhang +4
Documentation has long guided computer system tuning by distilling expert knowledge into per-parameter recommendations. Yet such guides capture only what experts conclude, discardi…
VUDA: Enabling Controlled Spatial Sharing of Graphics and Compute on NVIDIA GPUs
Bin Xu, Pengfei Hu, Wenxin Zheng +2
Graphics and compute increasingly share a single GPU in embodied AI simulators, AI-enabled games, and VR systems. Running these workloads concurrently can improve utilization, but…
SkVM: Revisiting Language VM for Skills across Heterogenous LLMs and Harnesses
Le Chen, Erhu Feng, Yubin Xia +1
LLM agents increasingly adopt skills as a reusable unit of composition. While skills are shared across diverse agent platforms, current systems treat them as raw context, causing t…
Determinism-Preserving GPU Spatial Sharing with Vitamin-E
Zhenyuan Yang, Wenxin Zheng, Mingyu Li +1
GPU sharing faces a determinism--utilization tradeoff: fixed bindings can strand capacity as demand fluctuates, while resource-driven kernel reshaping improves utilization by alter…
From Imperative to Declarative: Towards LLM-friendly OS Interfaces for Boosted Computer-Use Agents
Yuan Wang, Mingyu Li, Haibo Chen
Computer-use agents (CUAs) powered by large language models (LLMs) have emerged as a promising approach to automating computer tasks, yet they struggle with the existing human-orie…