[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"site-home-live-entry-setting":3,"site-header-contests":8,"site-header-camps":49,"study-material-detail-study-9":104},{"enabled":4,"title":5,"statusText":5,"buttonText":5,"campSlug":5,"stage":6,"classroomId":7,"hasClassroom":4},false,"",0,null,{"total":9,"list":10,"pageNum":44,"pageSize":45,"size":9,"startRow":44,"endRow":9,"pages":44,"prePage":6,"nextPage":6,"isFirstPage":46,"isLastPage":46,"hasPreviousPage":4,"hasNextPage":4,"navigatePages":47,"navigatepageNums":48,"navigateFirstPage":44,"navigateLastPage":44},3,[11,25,35],{"slug":12,"title":13,"logo":7,"liveUrl":14,"prize":15,"sponsor":16,"description":17,"organizer":18,"startTime":19,"endTime":20,"registrationDeadline":21,"status":22,"dashboardVisibility":23,"createdAt":24},"spring2026","2026春季启元人工智能大赛","https:\u002F\u002Fwww.bilibili.com\u002Fblackboard\u002Flive\u002Flive-activity-player.html?cid=1822703816&quality=0","200,000","麒麟软件有限公司","“2026春季启元人工智能大赛”是由启元实验室主办的面向全国人工智能领域开发者、科研人员及高校学生的权威赛事。大赛旨在促进与培养具备工程实践能力和创新能力的优秀人工智能领域开发者、推广国产软硬件技术与生态、助力国家人工智能产业高质量发展。\n各参赛选手需在规定的比赛周期内，依据赛事规则选定赛题并完成相应挑战，优胜者将共同角逐总额20万元的赛事奖金池，并且还将有机会获得实习和知名企业优质岗位内推资格。","启元实验室","2026-04-19 16:00:00","2026-07-13 16:00:00","2026-06-16 16:00:00","ongoing","private","2026-04-16 08:00:00",{"slug":26,"title":27,"logo":7,"liveUrl":14,"prize":28,"sponsor":7,"description":29,"organizer":18,"startTime":30,"endTime":31,"registrationDeadline":32,"status":33,"dashboardVisibility":23,"createdAt":34},"autumn2025","2025秋季启元人工智能大赛","150,000","“2025秋季启元人工智能大赛”是由启元实验室主办的面向全国人工智能领域开发者、科研人员及高校学生的权威赛事。大赛旨在促进与培养具备工程实践能力和创新能力的优秀人工智能领域开发者、推广国产软硬件技术与生态、助力国家人工智能产业高质量发展。","2025-11-08 08:00:00","2026-01-13 08:00:00","2025-12-30 08:00:00","ended","2025-11-03 08:00:00",{"slug":36,"title":37,"logo":7,"liveUrl":14,"prize":38,"sponsor":7,"description":39,"organizer":18,"startTime":40,"endTime":41,"registrationDeadline":42,"status":33,"dashboardVisibility":23,"createdAt":43},"summer2025","启元人工智能大赛(系统方向)","80,000","启元人工智能大赛由启元实验室主办，其中系统赛道基于启元实验室九源软件栈的相关开源项目，通过拉动开源社区力量，一方面促进国产智能计算生态的发展，另一方面也为人工智能系统方向的学生及从业者提供一个实践平台。","2025-07-08 08:00:00","2025-08-25 08:00:00","2025-07-22 08:00:00","2025-06-22 08:00:00",1,5,true,8,[44],{"total":45,"list":50,"pageNum":44,"pageSize":45,"size":45,"startRow":44,"endRow":45,"pages":44,"prePage":6,"nextPage":6,"isFirstPage":46,"isLastPage":46,"hasPreviousPage":4,"hasNextPage":4,"navigatePages":47,"navigatepageNums":103,"navigateFirstPage":44,"navigateLastPage":44},[51,62,73,84,94],{"id":52,"slug":53,"title":54,"logo":55,"publicLiveEmbedUrl":14,"description":56,"introduction":57,"organizer":18,"coOrganizer":5,"sponsor":16,"startTime":58,"endTime":59,"status":22,"dashboardVisibility":60,"createdAt":58,"updatedAt":61},"019eeeff-6b00-7d55-a629-efafded3e232","summer2026","2026年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2026-06\u002F019eef7f-7816-7828-9309-79c2c61dd568.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=o29Dy5KMbkQVCWTVaFUX1%2F9v%2FI8%3D","由启元实验室发起的InfiniTensor大模型与人工智能系统训练营，面向开源社区、高校及产业界伙伴开放。活动以「普及AI全栈知识技术、培养工程实践与创新型人才、推广国产软硬件技术生态」为核心目标，助力国家人工智能产业高质量发展。","...","2026-06-22 08:00:00","2026-10-19 08:00:00","public","2026-10-12 08:00:00",{"id":63,"slug":64,"title":65,"logo":66,"publicLiveEmbedUrl":7,"description":67,"introduction":57,"organizer":18,"coOrganizer":5,"sponsor":68,"startTime":69,"endTime":70,"status":33,"dashboardVisibility":23,"createdAt":71,"updatedAt":72},"019af3f3-e06a-7245-aa8d-37dc99a69c73","winter2025","2025年冬季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2025-12\u002F019afc69-b7ca-7805-91fb-605429671c01.png?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=RF6et0tKpeNsR%2FwxsCjy7zBTZbU%3D","由启元实验室发起的InfiniTensor大模型与人工智能系统训练营，面向开源社区、高校及产业界伙伴开放。活动以「普及AI全栈知识技术、培养工程实践与创新型人才、推广国产软硬件技术生态」为核心目标，助力国家人工智能产业高质量发展。 \n","新华三技术有限公司","2025-12-25 08:00:00","2026-03-27 08:00:00","2025-12-06 08:00:00","2026-10-22 20:24:30",{"id":74,"slug":36,"title":75,"logo":76,"publicLiveEmbedUrl":7,"description":77,"introduction":57,"organizer":18,"coOrganizer":5,"sponsor":78,"startTime":79,"endTime":80,"status":81,"dashboardVisibility":23,"createdAt":82,"updatedAt":83},"01980342-208c-7803-b3f3-19e1d31960ba","2025年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fuploads\u002F2025-07\u002F0198045e-28fe-78f0-95fe-23835ce58395.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=mArtxwAhCa0dE5Cm17SDDzrBtQ8%3D","InfiniTensor大模型与人工智能系统训练营由启元实验室及清华大学 AI 系统方向团队发起，面向开源社区、高校和产业界，普及人工智能各方面知识和技术，培养相关人才。训练营联合开源项目开发者、学术界大牛、工业界资深工程师参与授课与项目指导。课程采用「线上直播授课 + 在线答疑 + 项目实践辅导 」的方式进行，确保每位同学都能获得充分成长。训练营全程免费，通过视频、直播课、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。","北京积算科技有限公司","2025-07-14 08:00:00","2025-09-30 08:00:00","upcoming","2025-07-13 08:00:00","2026-10-22 20:24:58",{"id":85,"slug":86,"title":87,"logo":88,"publicLiveEmbedUrl":7,"description":89,"introduction":57,"organizer":18,"coOrganizer":5,"sponsor":5,"startTime":90,"endTime":91,"status":81,"dashboardVisibility":23,"createdAt":92,"updatedAt":93},"01980cf9-a7d8-7796-ae86-0bbe05334add","winter2024","2024年冬季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Ftesting-cdn.infinitensor.com\u002Fuploads\u002F2025-07\u002F01981269-ce23-7624-a287-9f07dd0d85f6.jpg","InfiniTensor大模型与人工智能系统训练营由清华大学发起，面向开源社区、高校和产业界，普及人工智能各方面知识和技术，培养相关人才。训练营全程免费，通过视频、直播课、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。报名学员请阅读“导学”课程页的训练营介绍。","2025-01-05 08:00:00","2025-03-22 08:00:00","2025-07-15 08:00:00","2026-10-22 20:24:39",{"id":95,"slug":96,"title":97,"logo":98,"publicLiveEmbedUrl":7,"description":99,"introduction":57,"organizer":18,"coOrganizer":5,"sponsor":5,"startTime":100,"endTime":101,"status":81,"dashboardVisibility":23,"createdAt":92,"updatedAt":102},"01980c20-363a-7dc3-8f5a-58a68e08dcc4","summer2024","2024年夏季InfiniTensor大模型与人工智能系统训练营","https:\u002F\u002Ftesting-cdn.infinitensor.com\u002Fuploads\u002F2025-07\u002F01981269-fbfa-788a-afcd-17fa9be74ff9.jpg","InfiniTensor大模型与人工智能系统训练营由清华大学发起，面向开源社区和高校，普及人工智能各方面知识和技术，培养相关人才。训练营全程免费，通过课程、习题和项目实践，在人工智能飞速发展的背景下，为学员提供一窥深层技术的机会。","2024-07-20 08:00:00","2024-09-29 08:00:00","2026-10-22 20:24:49",[44],{"currentItem":105,"processedData":125},{"img":106,"keywords":107,"treeType":44,"description":108,"showStatus":44,"updateTime":109,"type":110,"title":111,"content":112,"nextArticle":113,"createTime":109,"sortOrder":117,"canonicalPath":118,"id":119,"contentType":120,"prevArticle":121},"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_fm.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=pnC1hs2eDllB8lagfiOjVY5KUmw%3D","InfiniTensor,冬季训练营,AI编译器,模型加速,部署优化","介绍 AI 编译器概述与基础能力，帮助理解编译器在模型加速与部署中的核心作用。","2026-10-22 20:38:11",2,"InfiniTensor冬季训练营AI编译器方向第一课《AI编译器概述》","![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw1.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=H%2BAB3R9BSe1m9kfRWgW4G9Scb04%3D)\r\n\r\n\r\n2026 年 1 月 12 日，InfiniTensor 冬季大模型与人工智能系统训练营 AI 编译器方向第一课《AI 编译器概述》正式开讲。\r\n\r\n本节课主要讲述 AI 编译器的历史、架构、与传统编译器的区别，以及主流 AI 编译器（如 TVM 和 MLIR）的工作原理，并深入探讨 AI 程序中的核心概念——张量与计算图。\r\n\r\n### AI的历史与现状\r\n#### 1. AI的起源\r\n- 艾伦·图灵与图灵测试（1950年）\r\n- 1956年达特茅斯会议：AI的诞生与术语提出\r\n#### 2. AI的发展历程\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw2.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=p5SlYYoNP6VXscPcEWIpXXPJj4Y%3D)\r\n\r\n#### 3.AI系统架构\r\n- 应用层：CV、NLP、搜索推荐、大模型等\r\n- 开发框架层：Pytorch、Tensorflow 等AI框架，TensorRT 等推理引擎\r\n- 编程编译层：AI编译器核心职责与架构\r\n- AI芯片层：硬件接口、抽象、驱动与计算库\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw3.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=kleUeWCjMkxhoyA52qyWutXOlNI%3D)\r\n\r\n\r\n### 传统编译器与AI编译器对比\r\n#### 1.传统编译器发展\r\n- GCC：一对多架构，高耦合度，难以维护\r\n- LLVM：引入中间表示（IR），解耦前端与后端，提升模块化、可维护性与可复用性\r\n#### 2.传统编译器架构\r\n- 前端：词法分析、语法分析、语义分析，生成与硬件无关的中间表示\r\n- 优化器：围绕IR进行工作，通过pass进行独立分析或转换操作\r\n- 后端：将优化后的IR映射到具体目标硬件，进行指令选择、寄存器分配、指令调度\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw4.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=WVO6BPU7iGEtTY%2FAwjzStfk5WM4%3D)\r\n\r\n\r\n#### 3.AI编译器设计背景\r\n- 多样化深度学习框架与碎片化硬件计算平台\r\n#### 4.AI编译器与传统编译器的核心区别\r\n- 领域特定计算抽象优化能力\r\n- 适配异构与可配置AI加速器的能力\r\n- 区分训练与推理场景需求的能力\r\n#### 5.AI编译器目标\r\n- 接口层面：Python作为主要前端语言\r\n- 系统架构：多层中间表示设计\r\n- 优化能力：面向神经网络，集成训练与推理场景需求\r\n- 跨硬件代码生成能力\r\n\r\n### 主流AI编译器介绍\r\n#### 1.TVM（Tensor Virtual Machine）\r\n- 工作流程：\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw5.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=xItJV35Y%2BKaOEtgjiI4T%2F2OdnKw%3D)\r\n\r\n\r\n- 核心创新：解耦计算定义与执行调度，自动搜索最优代码实现方案\r\n#### 2.MLIR（Multilevel Intermediate Representation）\r\n- 核心思想：分层中间表示与可扩展方言系统\r\n- 解决的问题：编译器领域重复造轮子问题，通过统一IR与方言系统复用优化技术\r\n\r\n### AI程序的抽象：张量与计算图\r\n#### 1.张量（Tensor）\r\n- 定义：AI中对高维数组的统一称谓，动态化数据结构，封装数据类型、形状、名称与存储区域\r\n- 阶数与数学意义：0维标量、1维向量、2维矩阵、更高阶张量\r\n- 形状与语义：形状决定计算方式，如矩阵乘法\r\n- 存储布局：逻辑形状与物理存储解耦，引入步长与偏移概念，提升操作灵活性与执行效率\r\n- 变换操作：转置、切片、广播、重塑等\r\n\r\n#### 2.计算图（Computational Graph）\r\n- 定义：由拓扑、张量、算子三要素构成的AI程序表示\r\n- 拓扑：基于深度神经网络的有向无环图，表示数据流关系\r\n- 张量：操作大规模数据的高维数组\r\n- 算子：计算方式的表示\r\n- 图变化操作：张量推导、图优化、核函数选择与优化\r\n- ONNX（Open Neural Network Exchange）：开放格式，表示深度神经网络模型，提供模型互操作性基础\r\n- 张量推导：根据算子信息与输入张量信息推导输出张量信息，包括形状推导与数据类型推导\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw6.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=fqo24C3RD0YpIMg2jFVS1xTwqms%3D)\r\n\r\n\r\n### 总结与展望\r\n本节课了解了AI历史与现状，理解AI编译器的重要性；对比传统编译器与AI编译器，掌握AI编译器的核心优势；介绍了主流AI编译器TVM与MLIR，理解其工作原理与创新点；探讨AI程序中的核心概念——张量与计算图，掌握其基本操作与应用。后续将继续深入学习AI编译器相关技术。\r\n\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_01_zw6.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=nRtncQE8d9WeqAJkebtpCuy2cYI%3D)\r\n\r\n\r\n### 💡 学习贴士\r\n更多细节及完整课程内容，请观看直播或查看课程回放。\r\n\r\n链接：https:\u002F\u002Fwww.infinitensor.com\u002Fcamp\u002Fwinter2025\r\n\r\n### 📚 训练营报名持续进行中\r\n扫码报名与加入群聊，或点击 **「阅读原文」** 直达官网报名\r\n\r\n\r\n![](https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_01_zw7.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=cXo0Vj%2B8TkiETVdZDjCIcCGN6sw%3D)",{"publishTime":7,"createTime":109,"canonicalPath":114,"id":115,"title":116},"\u002Fstudy-material\u002Fstudy-8","study-8","InfiniTensor 冬季训练营推理引擎方向第一课《AI 模型导论与推理基础原理》",100,"\u002Fstudy-material\u002Fstudy-9","study-9","markdown",{"publishTime":7,"createTime":109,"canonicalPath":122,"id":123,"title":124},"\u002Fstudy-material\u002Fstudy-10","study-10","InfiniTensor冬季训练营CPU并行编程第一课《CPU性能优化导论》",{"htmlContent":126,"headings":127},"\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw1.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=H%2BAB3R9BSe1m9kfRWgW4G9Scb04%3D\" alt=\"\">\u003C\u002Fp>\n\u003Cp>2026 年 1 月 12 日，InfiniTensor 冬季大模型与人工智能系统训练营 AI 编译器方向第一课《AI 编译器概述》正式开讲。\u003C\u002Fp>\n\u003Cp>本节课主要讲述 AI 编译器的历史、架构、与传统编译器的区别，以及主流 AI 编译器（如 TVM 和 MLIR）的工作原理，并深入探讨 AI 程序中的核心概念——张量与计算图。\u003C\u002Fp>\n\u003Ch3 id=\"ai的历史与现状\">\u003Ca class=\"anchor\" href=\"#ai的历史与现状\" aria-label=\"跳转到标题：AI的历史与现状\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>AI的历史与现状\u003C\u002Fh3>\u003Ch4 id=\"1-ai的起源\">\u003Ca class=\"anchor\" href=\"#1-ai的起源\" aria-label=\"跳转到标题：1. AI的起源\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>1. AI的起源\u003C\u002Fh4>\u003Cul>\n\u003Cli>艾伦·图灵与图灵测试（1950年）\u003C\u002Fli>\n\u003Cli>1956年达特茅斯会议：AI的诞生与术语提出\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"2-ai的发展历程\">\u003Ca class=\"anchor\" href=\"#2-ai的发展历程\" aria-label=\"跳转到标题：2. AI的发展历程\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>2. AI的发展历程\u003C\u002Fh4>\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw2.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=p5SlYYoNP6VXscPcEWIpXXPJj4Y%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch4 id=\"3-ai系统架构\">\u003Ca class=\"anchor\" href=\"#3-ai系统架构\" aria-label=\"跳转到标题：3.AI系统架构\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>3.AI系统架构\u003C\u002Fh4>\u003Cul>\n\u003Cli>应用层：CV、NLP、搜索推荐、大模型等\u003C\u002Fli>\n\u003Cli>开发框架层：Pytorch、Tensorflow 等AI框架，TensorRT 等推理引擎\u003C\u002Fli>\n\u003Cli>编程编译层：AI编译器核心职责与架构\u003C\u002Fli>\n\u003Cli>AI芯片层：硬件接口、抽象、驱动与计算库\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw3.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=kleUeWCjMkxhoyA52qyWutXOlNI%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch3 id=\"传统编译器与ai编译器对比\">\u003Ca class=\"anchor\" href=\"#传统编译器与ai编译器对比\" aria-label=\"跳转到标题：传统编译器与AI编译器对比\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>传统编译器与AI编译器对比\u003C\u002Fh3>\u003Ch4 id=\"1-传统编译器发展\">\u003Ca class=\"anchor\" href=\"#1-传统编译器发展\" aria-label=\"跳转到标题：1.传统编译器发展\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>1.传统编译器发展\u003C\u002Fh4>\u003Cul>\n\u003Cli>GCC：一对多架构，高耦合度，难以维护\u003C\u002Fli>\n\u003Cli>LLVM：引入中间表示（IR），解耦前端与后端，提升模块化、可维护性与可复用性\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"2-传统编译器架构\">\u003Ca class=\"anchor\" href=\"#2-传统编译器架构\" aria-label=\"跳转到标题：2.传统编译器架构\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>2.传统编译器架构\u003C\u002Fh4>\u003Cul>\n\u003Cli>前端：词法分析、语法分析、语义分析，生成与硬件无关的中间表示\u003C\u002Fli>\n\u003Cli>优化器：围绕IR进行工作，通过pass进行独立分析或转换操作\u003C\u002Fli>\n\u003Cli>后端：将优化后的IR映射到具体目标硬件，进行指令选择、寄存器分配、指令调度\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw4.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=WVO6BPU7iGEtTY%2FAwjzStfk5WM4%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch4 id=\"3-ai编译器设计背景\">\u003Ca class=\"anchor\" href=\"#3-ai编译器设计背景\" aria-label=\"跳转到标题：3.AI编译器设计背景\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>3.AI编译器设计背景\u003C\u002Fh4>\u003Cul>\n\u003Cli>多样化深度学习框架与碎片化硬件计算平台\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"4-ai编译器与传统编译器的核心区别\">\u003Ca class=\"anchor\" href=\"#4-ai编译器与传统编译器的核心区别\" aria-label=\"跳转到标题：4.AI编译器与传统编译器的核心区别\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>4.AI编译器与传统编译器的核心区别\u003C\u002Fh4>\u003Cul>\n\u003Cli>领域特定计算抽象优化能力\u003C\u002Fli>\n\u003Cli>适配异构与可配置AI加速器的能力\u003C\u002Fli>\n\u003Cli>区分训练与推理场景需求的能力\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"5-ai编译器目标\">\u003Ca class=\"anchor\" href=\"#5-ai编译器目标\" aria-label=\"跳转到标题：5.AI编译器目标\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>5.AI编译器目标\u003C\u002Fh4>\u003Cul>\n\u003Cli>接口层面：Python作为主要前端语言\u003C\u002Fli>\n\u003Cli>系统架构：多层中间表示设计\u003C\u002Fli>\n\u003Cli>优化能力：面向神经网络，集成训练与推理场景需求\u003C\u002Fli>\n\u003Cli>跨硬件代码生成能力\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3 id=\"主流ai编译器介绍\">\u003Ca class=\"anchor\" href=\"#主流ai编译器介绍\" aria-label=\"跳转到标题：主流AI编译器介绍\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>主流AI编译器介绍\u003C\u002Fh3>\u003Ch4 id=\"1-tvm-tensor-virtual-machine\">\u003Ca class=\"anchor\" href=\"#1-tvm-tensor-virtual-machine\" aria-label=\"跳转到标题：1.TVM（Tensor Virtual Machine）\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>1.TVM（Tensor Virtual Machine）\u003C\u002Fh4>\u003Cul>\n\u003Cli>工作流程：\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw5.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=xItJV35Y%2BKaOEtgjiI4T%2F2OdnKw%3D\" alt=\"\">\u003C\u002Fp>\n\u003Cul>\n\u003Cli>核心创新：解耦计算定义与执行调度，自动搜索最优代码实现方案\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"2-mlir-multilevel-intermediate-representation\">\u003Ca class=\"anchor\" href=\"#2-mlir-multilevel-intermediate-representation\" aria-label=\"跳转到标题：2.MLIR（Multilevel Intermediate Representation）\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>2.MLIR（Multilevel Intermediate Representation）\u003C\u002Fh4>\u003Cul>\n\u003Cli>核心思想：分层中间表示与可扩展方言系统\u003C\u002Fli>\n\u003Cli>解决的问题：编译器领域重复造轮子问题，通过统一IR与方言系统复用优化技术\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch3 id=\"ai程序的抽象-张量与计算图\">\u003Ca class=\"anchor\" href=\"#ai程序的抽象-张量与计算图\" aria-label=\"跳转到标题：AI程序的抽象：张量与计算图\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>AI程序的抽象：张量与计算图\u003C\u002Fh3>\u003Ch4 id=\"1-张量-tensor\">\u003Ca class=\"anchor\" href=\"#1-张量-tensor\" aria-label=\"跳转到标题：1.张量（Tensor）\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>1.张量（Tensor）\u003C\u002Fh4>\u003Cul>\n\u003Cli>定义：AI中对高维数组的统一称谓，动态化数据结构，封装数据类型、形状、名称与存储区域\u003C\u002Fli>\n\u003Cli>阶数与数学意义：0维标量、1维向量、2维矩阵、更高阶张量\u003C\u002Fli>\n\u003Cli>形状与语义：形状决定计算方式，如矩阵乘法\u003C\u002Fli>\n\u003Cli>存储布局：逻辑形状与物理存储解耦，引入步长与偏移概念，提升操作灵活性与执行效率\u003C\u002Fli>\n\u003Cli>变换操作：转置、切片、广播、重塑等\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch4 id=\"2-计算图-computational-graph\">\u003Ca class=\"anchor\" href=\"#2-计算图-computational-graph\" aria-label=\"跳转到标题：2.计算图（Computational Graph）\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>2.计算图（Computational Graph）\u003C\u002Fh4>\u003Cul>\n\u003Cli>定义：由拓扑、张量、算子三要素构成的AI程序表示\u003C\u002Fli>\n\u003Cli>拓扑：基于深度神经网络的有向无环图，表示数据流关系\u003C\u002Fli>\n\u003Cli>张量：操作大规模数据的高维数组\u003C\u002Fli>\n\u003Cli>算子：计算方式的表示\u003C\u002Fli>\n\u003Cli>图变化操作：张量推导、图优化、核函数选择与优化\u003C\u002Fli>\n\u003Cli>ONNX（Open Neural Network Exchange）：开放格式，表示深度神经网络模型，提供模型互操作性基础\u003C\u002Fli>\n\u003Cli>张量推导：根据算子信息与输入张量信息推导输出张量信息，包括形状推导与数据类型推导\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_09_zw6.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=fqo24C3RD0YpIMg2jFVS1xTwqms%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch3 id=\"总结与展望\">\u003Ca class=\"anchor\" href=\"#总结与展望\" aria-label=\"跳转到标题：总结与展望\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>总结与展望\u003C\u002Fh3>\u003Cp>本节课了解了AI历史与现状，理解AI编译器的重要性；对比传统编译器与AI编译器，掌握AI编译器的核心优势；介绍了主流AI编译器TVM与MLIR，理解其工作原理与创新点；探讨AI程序中的核心概念——张量与计算图，掌握其基本操作与应用。后续将继续深入学习AI编译器相关技术。\u003C\u002Fp>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_01_zw6.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=nRtncQE8d9WeqAJkebtpCuy2cYI%3D\" alt=\"\">\u003C\u002Fp>\n\u003Ch3 id=\"学习贴士\">\u003Ca class=\"anchor\" href=\"#学习贴士\" aria-label=\"跳转到标题：💡 学习贴士\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>💡 学习贴士\u003C\u002Fh3>\u003Cp>更多细节及完整课程内容，请观看直播或查看课程回放。\u003C\u002Fp>\n\u003Cp>链接：\u003Ca href=\"https:\u002F\u002Fwww.infinitensor.com\u002Fcamp\u002Fwinter2025\">https:\u002F\u002Fwww.infinitensor.com\u002Fcamp\u002Fwinter2025\u003C\u002Fa>\u003C\u002Fp>\n\u003Ch3 id=\"训练营报名持续进行中\">\u003Ca class=\"anchor\" href=\"#训练营报名持续进行中\" aria-label=\"跳转到标题：📚 训练营报名持续进行中\">\u003Cspan class=\"octicon octicon-link\">\u003C\u002Fspan>\u003C\u002Fa>📚 训练营报名持续进行中\u003C\u002Fh3>\u003Cp>扫码报名与加入群聊，或点击 \u003Cstrong>「阅读原文」\u003C\u002Fstrong> 直达官网报名\u003C\u002Fp>\n\u003Cp>\u003Cimg src=\"https:\u002F\u002Foss.infinitensor.com\u002Fprod\u002Fimages\u002FKnowledge_Base\u002FKB_01_zw7.jpg?Expires=1792752820&OSSAccessKeyId=LTAI5t9R6uzvTa69CT92cFvV&Signature=cXo0Vj%2B8TkiETVdZDjCIcCGN6sw%3D\" alt=\"\">\u003C\u002Fp>\n",[128,131,135,138,141,144,147,150,153,156,159,162,165,168,171,174,177,179,182],{"id":129,"text":130,"level":9},"ai的历史与现状","AI的历史与现状",{"id":132,"text":133,"level":134},"1-ai的起源","1. AI的起源",4,{"id":136,"text":137,"level":134},"2-ai的发展历程","2. AI的发展历程",{"id":139,"text":140,"level":134},"3-ai系统架构","3.AI系统架构",{"id":142,"text":143,"level":9},"传统编译器与ai编译器对比","传统编译器与AI编译器对比",{"id":145,"text":146,"level":134},"1-传统编译器发展","1.传统编译器发展",{"id":148,"text":149,"level":134},"2-传统编译器架构","2.传统编译器架构",{"id":151,"text":152,"level":134},"3-ai编译器设计背景","3.AI编译器设计背景",{"id":154,"text":155,"level":134},"4-ai编译器与传统编译器的核心区别","4.AI编译器与传统编译器的核心区别",{"id":157,"text":158,"level":134},"5-ai编译器目标","5.AI编译器目标",{"id":160,"text":161,"level":9},"主流ai编译器介绍","主流AI编译器介绍",{"id":163,"text":164,"level":134},"1-tvm-tensor-virtual-machine","1.TVM（Tensor Virtual Machine）",{"id":166,"text":167,"level":134},"2-mlir-multilevel-intermediate-representation","2.MLIR（Multilevel Intermediate Representation）",{"id":169,"text":170,"level":9},"ai程序的抽象-张量与计算图","AI程序的抽象：张量与计算图",{"id":172,"text":173,"level":134},"1-张量-tensor","1.张量（Tensor）",{"id":175,"text":176,"level":134},"2-计算图-computational-graph","2.计算图（Computational Graph）",{"id":178,"text":178,"level":9},"总结与展望",{"id":180,"text":181,"level":9},"学习贴士","💡 学习贴士",{"id":183,"text":184,"level":9},"训练营报名持续进行中","📚 训练营报名持续进行中"]