当前位置:首页 > 报告详情

利用智能体人工智能加速固件开发和部署.pdf

上传人: 明**** 编号:1011787 2025-12-21 18页 1.29MB

1、Chitrak Gupta(AMI)Sudan Varadachari Ayanam(AMI)Turbocharging Firmware Development and Deployment using Agentic AITurbocharging Firmware Development and Deployment using Agentic AI Chitrak GuptaSudan Varadachari AyanamOCP SPECIAL FOCUS:ARTIFICIAL INTELLIGENCE(AI)The rapid evolution of server technolo

2、gies and firmware complexity demands a shift from manual coding to intelligent automation.This session presents a novel framework using Generative AI to streamline platform code generation for server firmware orchestration.It enables automated creation of platform porting files,Redfish schemas for O

3、penBMC,and UEFI debug log analysis from simple prompts.Integrated test case agents validate generated code against system specs.Real-world examples demonstrate how Agentic AI drives efficiency,accelerates open standard adoption,and reduces time to market for ODM/OEM platforms.Abstract:AMI AI-Driven

4、Build Orchestration for DC-MHS Plug-n-PlayDC-MHS standardizes interoperable,modular server building blocks(HPMs,DC-SCM,power,I/O,chassis),enabling mix-and-match ecosystems across vendors and form factorsThe DC-MHS PnP forum is defining platform identification files to facilitate plug-n-play and inte

5、roperability(Modular-Plug-n-Play FRU,Discovery&Boot Design Specification)The M-PnP workstream(Modular Plug-n-Play)focuses on discovery,identification,and“just-works”integration of modulesso a platform reconfiguration can be recognized and brought up with minimal manual effort.AMI proposes an automat

6、ed firmware management system that enables platform code generation,redfish specification updates,automated test pipeline and issue debugging for OpenBMC and Uefi based stacksIntroductionAMI Firmware Management System Embed AI Enabled Operations,inside the firmware orchestration pipeline so platform

word格式文档无特别注明外均可编辑修改,预览文件经过压缩,下载原文更清晰!
三个皮匠报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
根据报告的内容,全文主要内容概括如下: - **AI在服务器固件开发中的应用**:使用Agentic AI框架自动化生成服务器固件代码,提高效率,加速开放标准的采用,缩短上市时间。 - **DC-MHS标准**:DC-MHS标准化模块化服务器构建块,实现跨供应商和形态的混合匹配生态系统。 - **AMI固件管理系统**:集成AI操作,自动触发平台文件生成、代码修改和测试,减少手动移植周期。 - **Agentic AI流程**:检测平台变化,触发相应的代理处理模块描述文件和配置文件,生成或修改Redfish代码,并执行自动化测试。 - **关键数据**: - 平台移植效率提高约15%,每周节省每位工程师6小时。 - 问题解决时间减少约25%。 - 测试周期缩短20%。 - **AMI Uefi解决方案**:AI代理提供调试问题机制,分析原始BIOS日志和BMC日志,提供代码修复建议。 - **合作与贡献**:AMI欢迎硅和设备制造合作伙伴在固件管理空间合作,并参与AI基础设施的开放系统。
"AI加速服务器固件开发?" "自动化测试,AI如何助力?" "AI在服务器管理中的突破!"
客服
商务合作
小程序
服务号
折叠