English(EN) Atlas: Optimizing Deployment of Compound AI Workflows on Heterogeneous Clusters

Atlas 框架优化异构集群上的 AI 工作流部署

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-07 04:00

研究人员开发了 Atlas，一个旨在优化复杂 AI 工作流在异构集群上部署的框架。Atlas 解决了复合 AI 工作流中准确性估计的挑战，因为错误会在不同阶段之间传播。它引入了马尔可夫准确性预测器 (MAP)，通过离散化中间输出和组合转换配置文件来估计配置准确性，从而避免了昂贵的端到端分析。这种方法使 Atlas 能够将执行计划选择制定为混合整数线性规划问题，在满足服务水平目标 (SLO) 的同时最大化预测准确性并降低部署成本。 AI

影响该框架通过改进准确性估计和资源分配，有望实现更高效、更具成本效益的复杂 AI 系统的部署。

排序理由这是一篇详细介绍用于优化 AI 工作流部署的新框架和方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 32 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 这是一篇详细介绍用于优化 AI 工作流部署的新框架和方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: infra, paper; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.AI TIER_1 English(EN) · Milos Gravara, Andrija Stanisic, Stefan Nastic · 2026-09-07 04:00

Atlas：在异构集群上优化复合AI工作流的部署

arXiv:2609.04513v1 Announce Type: cross Abstract: Compound AI workflows are increasingly used to serve complex AI tasks by coordinating multiple AI models and software components. This approach enables deployment flexibility, as each workflow stage can expose different model vari…

本文如何被排名

报道来源 [1]

Atlas：在异构集群上优化复合AI工作流的部署

相关实体

相关话题