首页 > 科技 > 计算机 > 计算机科学

机密虚拟化：云计算大模型时代的数据安全新范式

作译者：宋川,朱运阁

出版时间：2025-08 千字数：353 版次：01-01 页数：252

开本：16开装帧： I S B N ：9787121511066

换版：

纸质书定价：￥108.0

库存:无

分享到：

共有图书评论0条【查看评论摘要】

看了又看

{{data.bookName}}

￥{{data.price}}

内容简介

前言

上架建议

作者简介

编辑推荐

音视频专区

目录
第 1 章 数据安全与隐私保护····························································.3
1.1 数字化发展带来的机遇·································································.4
1.1.1 数字化的价值·····································································.4 
1.1.2 国内数字化发展战略····························································.5 
1.1.3 全球各国数字化发展战略······················································.6 
1.2 数据安全面临的挑战····································································.7
1.2.1 数据隐私保护的重要性·························································.7 
1.2.2 数据安全的影响··································································.8 
1.2.3 数据生命周期的安全····························································.9 
1.3 隐私保护技术············································································.11
1.3.1 隐私保护技术基本概念························································.11 
1.3.2 隐私保护技术基本分类························································.12 
1.3.3 各类技术比较····································································.17 
第 2 章 云计算中的机密计算···························································.19
2.1 云计算及数据安全需求································································.20
2.1.1 云部署下的数据安全···························································.20
2.1.2 数据全生命周期保护····························································21 
2.2 机密计算技术演进 ·······································································22
2.2.1 机密计算的定义··································································22 
2.2.2 发展历程及现状··································································23 
第二篇 架构实现 
第 3 章 机密计算技术·······································································29
3.1 可信执行环境技术产生 ·································································30
3.2 防护域和攻击模型 ·······································································30
3.3 机密计算和可信执行环境技术纵览 ··················································31
3.3.1 ARM 架构 ·········································································32 
3.3.2 x86 架构············································································36 
3.3.3 RISC-V 架构 ······································································42 
3.3.4 特性差异···········································································46 
3.4 机密虚拟化················································································47
第 4 章 机密虚拟化架构与实现 ························································48
4.1 微架构······················································································49
4.1.1 威胁模型···········································································50 
4.1.2 架构设计···········································································51 
4.1.3 TCB 构成 ··········································································53 
4.1.4 内存保护机制·····································································54 
4.2 指令体系 ···················································································54
4.2.1 指令体系···········································································55 
4.2.2 元数据管理········································································57 
4.2.3 内存管理···········································································60 
4.2.4 处理器虚拟化·····································································62 
4.2.5 服务型可信域·····································································66
4.2.6 度量与认证·······································································.66 
4.3 虚拟化软件···············································································.68
4.3.1 虚拟化原理·······································································.68 
4.3.2 虚拟机软件的实现······························································.69 
4.3.3 虚拟机监控器实现······························································.72 
4.4 I/O 设备虚拟化··········································································.77
4.4.1 传统 I/O 设备 ····································································.77 
4.4.2 TEE-I/O 设备 ····································································.79 
4.4.3 TEE-I/O 安全模型 ······························································.81 
4.4.4 TEE-I/O 设备认证 ······························································.84 
第 5 章 高级特性探秘 ·····································································.89
5.1 远程认证··················································································.90
5.1.1 可信域度量信息生成···························································.91 
5.1.2 可信域引证生成·································································.93 
5.1.3 度量报告及生成·································································.94 
5.1.4 可信域引证数据结构···························································.98 
5.1.5 可信域引证验证·································································100 
5.2 热迁移·····················································································102
5.2.1 热迁移流程·······································································103 
5.2.2 状态和数据迁移·································································107 
5.3 嵌套虚拟化···············································································108
5.4 TCB 在线升级···········································································108
5.5 内存完整性···············································································109
第 6 章 机密虚拟化软件形态···························································115
6.1 机密虚拟机···············································································116
6.1.1 虚拟化技术原理·································································116 
6.1.2 机密虚拟机技术概念及发展··················································117
6.1.3 安全机制········································································.118 
6.1.4 I/O 数据保护 ···································································.120 
6.2 机密容器 ················································································.121
6.2.1 容器运行时安全·······························································.121 
6.2.2 机密容器架构··································································.123 
6.2.3 主要特性········································································.124 
6.3 安全操作系统 ··········································································.129
6.3.1 操作系统安全··································································.129 
6.3.2 星绽操作系统内核····························································.132 
6.3.3 基于机密计算构建安全操作系统 ··········································.133 
6.4 TDX 的系统软件栈···································································.138
6.4.1 基本组件········································································.138 
6.4.2 Linux 发行版的支持 ··························································.140 
第三篇 实践案例 
第 7 章 联邦学习···········································································.145
7.1 联邦学习介绍 ··········································································.146
7.2 机密计算与联邦学习的结合 ························································.148
7.3 横向联邦学习方案 ····································································.149
第 8 章 可信大模型·······································································.153
8.1 构建安全可信大模型 ·································································.154
8.1.1 大模型数据安全隐患·························································.154 
8.1.2 机密计算助力构建可信大模型 ·············································.156 
8.2 可信大模型应用场景 ·································································.157
8.3 大模型密态计算平台案例 ···························································.159
8.3.1 TrustFlow ·······································································.160
8.3.2 蚂蚁密算大模型服务···························································163 
第 9 章 云数据库 ············································································167
9.1 云数据库与数据安全···································································168
9.2 全密态数据库············································································169
9.3 典型案例··················································································173
9.3.1 瑶池全密态数据库······························································173 
9.3.2 EdgelessDB ·······································································174 
9.3.3 高斯密态数据库·································································175 
第 10 章 区块链··············································································177
10.1 区块链技术·············································································178
10.2 区块链应用的挑战 ····································································179
10.3 典型案例················································································181
10.3.1 Azure 机密账本 ································································181 
10.3.2 蚂蚁隐私保护合约链 ·························································182 
10.3.3 机密计算在隐私公链中的应用 ·············································184 
第 11 章 异构计算 ··········································································187
11.1 异构计算与安全性挑战 ······························································188
11.1.1 异构计算 ········································································188 
11.1.2 优势分析 ········································································189 
11.1.3 安全性挑战 ·····································································189 
11.2 异构机密计算··········································································191
11.2.1 发展历程 ········································································192 
11.2.2 商用机密计算 GPU····························································193 
11.3 应用案例················································································195
11.3.1 异构计算中的远程认证 ······················································195 
11.3.2 构建机密 AI 训练······························································197
第 12 章 远程认证服务 ·································································.199
12.1 MAA ···················································································.200
12.1.1 MAA 概览·····································································.200 
12.1.2 MAA 应用案例·······························································.201 
12.2 ITA ·····················································································.202
12.2.1 ITA 架构·······································································.202 
12.2.2 ITA 应用案例·································································.205 
第四篇 未来展望 
第 13 章 安全防护的持续完善·······················································.213
13.1 侧信道防御能力提升 ·······························································.214
13.2 可信性的增强 ········································································.215
13.2.1 主要局限 ······································································.215 
13.2.2 发展方向 ······································································.216 
13.2.3 零知识证明应用 ·····························································.217 
13.3 异构计算的协同保护 ·······························································.218
第 14 章 生态系统的协同发展·······················································.221
14.1 法规与监管体系 ·····································································.222
14.1.1 隐私保护立法 ································································.222 
14.1.2 跨境数据流动体系建设 ····················································.223 
14.2 多元技术融合 ········································································.224
14.2.1 隐私计算融合 ································································.224 
14.2.2 软件供应链安全 ·····························································.225 
14.3 标准化生态 ···········································································.226
参考文献 ·························································································.228

推荐序
推荐序1
在数字经济时代，数据已成为核心生产要素，但其流动性与共享需求也面临前所未有的挑战。数据安全与隐私保护是数字经济健康发展的基石，如何在充分发挥数据价值的同时保障数据安全，成为亟待解决的重要课题。
机密计算通过构建基于硬件的可信执行环境，实现对数据在使用过程中的保护，为解决上述问题提供了创新方案。机密虚拟化技术融合了机密计算与虚拟化技术，在保持虚拟化灵活性和兼容性的基础上，为虚拟机环境提供了强大的安全保障。相较于进程级机密计算，机密虚拟化显著降低了应用门槛。在严谨安全设计的保障下，基于机密虚拟化的数据要素流通实践将加速涌现。
本书系统且全面地介绍了机密虚拟化技术的理论基础、架构实现、实践应用和未来发展。作者从基础概念入手，深入剖析了以英特尔TDX为代表的机密虚拟化技术架构，并通过联邦学习、可信大模型、机密数据库等典型应用场景，展示了机密虚拟化技术的实际价值。
在数据要素市场化进程加速的背景下，本书的出版恰逢其时。相较于机密计算领域已出版书籍，本书独树一帜，通过理论与实践的结合、多技术路线的并举、前沿性与实用性的兼具，为读者提供了一本全面、深入、实用的技术指南。
正如加密技术重塑了互联网的信任机制，以机密计算为代表的密态计算技术，将重塑数据安全使用的规则，引领我们进入密态时代。本书正是理解并驾驭这次时代变革的必备指南，推荐所有关心数据安全、系统安全、数据要素流通的同仁阅读。

闫守孟 
蚂蚁密算科技CTO 


推荐序2
随着企业数字化进程的加速，企业 IT 系统上云已经成为广大企业的首选。云计算技术的资源弹性灵活和开箱即用的特点，满足了企业IT系统建设的需求。同时，云计算多租户的技术属性也让人们更重视云上数据的安全，而大模型技术的广泛应用，进一步引起了让人们对训练数据和大模型安全的重视。
然而，数据和信息安全技术覆盖范围广泛，芯片计算和数据安全涉及芯片底层的技术原理，往往难以理解。同时，市面上缺少对机密计算技术尤其是最新的机密虚拟化技术进行系统化介绍的文献或者书籍。
英特尔作为计算机芯片技术的开拓者，在CPU芯片和基础架构安全等领域有着深厚的技术积累。本书作者均来自英特尔，在机密计算、虚拟化和云计算领域深耕多年，具备丰富的实战经验和扎实的技术积累。本书聚焦最新的机密虚拟化技术，围绕其技术原理、工程实现与应用路径，进行了全面和深入的分析，能够帮助大家理解并合理运用该项技术。
本书首先介绍了数据安全和机密计算的发展历程和基本概念，然后从芯片指令体系、虚拟化软件、I/O设备虚拟化等方面详细讲述机密虚拟化技术的技术架构，最后通过联邦学习、可信大模型、云数据库、区块链、异构计算、远程认证服务等几个方面介绍了机密计算技术的实践案例和应用场景。 
本书可以帮助IT技术人员及云计算从业者理解机密计算技术的核心原理，保障企业 IT 系统和数据的云上安全。

郑然 
百度杰出架构师，百度智能云技术委员会联席主席


推荐序3
数智化技术的快速发展与广泛应用显著加快了社会创新的步伐，并深刻引领着产业变革的浪潮。数智化系统及数据资源的重要性急剧增加，成为不可或缺的核心要素。与此同时，其价值与风险并存，从个人隐私到企业机密，从国家安全到数字经济的健康发展，数据安全与隐私保护已不再是一个技术命题，而是关乎社会信任与科技发展的核心议题。然而，传统安全架构在应对AI大模型、异构计算等新兴场景时逐渐显露出局限性，亟须一种全新的范式来重构安全边界。
作为一本兼具深度与广度的技术著作，本书以机密计算技术为核心主线，构建了一个完整的知识体系。从基础概念到技术内核，作者以严谨的工程和全栈安全视角还原了机密计算的实现路径，展现了硬件虚拟化与安全隔离的深度融合。同时，结合联邦学习、可信大模型、区块链等前沿场景，通过对阿里云、蚂蚁、英伟达等真实产品案例的介绍，为行业提供了重要的实践价值和参考范式。此外，本书展望了机密计算技术演进的路线，强调了标准化生态与法规监管的重要性，体现了对技术发展与治理平衡的思考。
本书不仅是一本技术手册，更是一份推动行业生态协同的指南。在数据安全成为全球竞争高地的今天，本书的出版恰逢其时，它不仅回应了时代的迫切需求，更为技术革新与产业变革提供了坚实的理论支撑与实践指引。

龙勤
阿里云基础软件系统安全团队负责人，龙蜥社区安全联盟主席


推荐序4 
数据已经成为数字时代的基石！过去十年，全球数据量以约25%的年复合增长率迅猛增长，预计到2025年将达到惊人的181ZB。传统的数据安全手段主要聚焦于数据在传输和存储过程中的保护，但在使用和处理阶段，数据的安全依然存在明显的短板。 机密计算正是为此而生，它通过基于硬件的可信执行环境，在数据使用和处理过程中实现持续保护。在云部署时代，数据使用和处理阶段往往也是面临风险最大的阶段。
虚拟化技术是支撑云计算和大模型的基础计算架构，本书深入探讨了机密计算与其的融合方法。
本书第一部分介绍了数据隐私的基本原理，概述了数字时代面临的安全挑战，并系统梳理了各类隐私保护技术。第二部分介绍了多种机密计算架构，随后以英特尔的信任域扩展技术为例深入剖析了机密虚拟化技术的实现原理。
尤为重要的是，本书并未局限于理论探讨，第三部分着力将理论转化为实践，带领读者深入探索机密计算以及机密虚拟化技术在联邦学习、云数据库、大语言模型、区块链及异构计算场景中的实际应用。这些案例不仅彰显了机密虚拟化技术的巨大潜力，亦为从业者、研究人员及企业实施同类解决方案提供了实用性的操作指南。
第四部分着眼未来，探讨攻击者与防御者间持续的攻防博弈，剖析生态系统标准化的必要性，并阐释政策法规在构建可信数字基础设施过程中起到的关键作用。本书借此提示读者：技术并非万能的，信任需要系统性构建、持续性维护与科学性治理。
本书是技术人员、云架构师、网络安全从业者及政策制定者的必备工具书。无论您是从事安全平台设计、隐私技术研究，还是制定数字政策，都能从中获得兼具技术深度与战略前瞻的专业指导。
信任、隐私与安全终将定义人类的数字未来。作者通过这本著作，为构建可信安全的数字未来提供了一份路线图与实践工具指南。

Mohan J. Kumar
英特尔院士

 
推荐序5
机密计算范式是过去十年计算机安全领域中最令人兴奋的发展之一。它推动了进攻性和防御性安全研究探索，在数据和信息的价值成为影响当今世界的关键因素，并在人工智能驱动的未来愈发重要的背景下，这些探索变得日益重要。
英特尔在开放平台的安全技术研发上有着引以为傲的历史，这些技术最终推动了计算范式向机密计算的演进。今天的机密计算技术可以追溯到英特尔可信执行技术和可信平台模块，它们首次引入了认证原则以及对软件系统进行度量的能力。从那时起，英特尔推出了软件保护扩展技术，为应用程序提供了强大的机密计算安全隔离能力，而最近推出的可信域扩展技术，将数据安全隔离能力延伸至整个虚拟机。
机密计算范式开启了全新的数据安全工具时代，赋能全球信息经济持续发展。之所以称其为“范式”，是因为迈向机密计算不仅需要硬件的变革（处理器和关键计算设备），还需要对其上运行的软件栈（操作系统、虚拟机管理器、应用程序等）进行调整，并引入通过认证来满足平台可信度的需求。
本书作者团队长期致力于机密计算技术的研究与实践，并积极将其应用于解决当今乃至未来的关键问题。这本书凝聚了他们在多年使用机密计算技术解决现实问题过程中积累的实践经验和理论知识。
本书分为四部分：第一部分提供理解机密计算范式所需的基本概念和背景；第二部讲解机密虚拟化所需的硬件和软件架构实现；第三部分展示机密计算以及机密虚拟化令人兴奋的实际应用；第四部分展望机密计算的发展前景。
相信任何对机密计算感兴趣的从业者和研究者都能从本书中得到有益的收获。

Simon Johnson
英特尔院士，机密计算架构负责人

专家寄语1
大模型的快速发展加速了云计算的普及，也让数据安全和隐私保护面临前所未有的挑战。本书凝聚了作者在机密计算领域的深厚造诣，系统梳理了从机密虚拟化的架构设计到应用落地的技术全景。内容兼具理论深度与实践价值，为构建可信基础设施提供了宝贵参考，是相关领域从业者和研究者深入了解这个领域的重要读物。

刘煜堃
阿里云安全保障研发负责人


专家寄语2
在大模型引领的数字化浪潮中，数据安全与隐私保护已成为技术发展的核心命题。本书以机密计算为技术支点，系统解析虚拟化场景中可信执行环境的底层架构，覆盖了机密大模型推理等前沿场景的案例，并探讨了异构计算协同防护等技术趋势，为云计算与人工智能安全领域的从业者及研究者构建了从理论到产业落地的知识体系，非常值得一读！

张尧
抖音集团机密计算技术专家


专家寄语3
机密计算作为一项融合体系架构、操作系统、密码学等多元领域的创新安全技术，其复杂性使得深入理解颇具挑战。本书难能可贵地以“垂直整合”的视角，系统梳理了机密虚拟化的核心技术栈，为读者构建起全面且深入的认知框架。它不仅有助于突破技术门槛，更能启发业界在数据安全与隐私保护领域的新思考，是理论与实践兼备的扛鼎之作。
田洪亮
蚂蚁集团技术研究院研究员


专家寄语4
在云计算与大模型技术迅猛发展的今天，数据安全与隐私保护已成为全球关注的焦点。本书的问世恰逢其时，为这个领域提供了系统和深入的理论指导与实践参考。本书从基础概念到架构实现，从技术原理到实践案例，全面剖析了机密虚拟化的技术原理与应用场景，不仅深入探讨了数据安全、隐私保护和机密计算等关键技术，还结合联邦学习、可信大模型和云数据库等前沿应用，介绍了机密虚拟化的广阔前景。书中对异构计算、远程认证服务等新兴领域的探索，更是为读者打开了未来技术发展的新视野。
作为一项颠覆性的技术，机密虚拟化正在重塑云计算与大模型时代的安全范式。本书的出版，不仅为从业者提供了宝贵的知识资源，也为学术界和产业界的进一步研究与实践奠定了坚实基础。相信本书将成为推动数据安全技术发展的重要力量，助力构建更加安全、可信的数字未来。衷心祝愿本书的读者能够从中汲取灵感，共同探索机密虚拟化技术的无限可能，为数字时代的安全与创新贡献力量。
张锋巍
南方科技大学研究生院副院长，长聘副教授、研究员、博导、国家级青年人才
前言
随着数据安全和隐私保护需求的不断提升，以实现“数据可用不可见”为核心目标的隐私计算理念逐渐深入人心。其中，基于硬件辅助的机密计算技术凭借卓越的安全性与高效性，在数据隐私保护与计算效率之间取得良好平衡，成为隐私计算领域“鱼与熊掌可兼得”的主线技术方案。
今天的云计算和数据中心主要构建在虚拟化架构之上，虚拟化技术提供了资源隔离、弹性扩展等多种技术优势，实现计算资源的高效分配和灵活调度。然而，虚拟化技术诞生伊始主要侧重于提升资源利用率和系统性能，而数据安全和隐私保护，尤其是“数据可用而不见”的理念，并不是传统的虚拟化架构设计时的核心考量。伴随着云计算的应用以及数据价值的不断提升，其在数据安全及隐私保护等方面面临诸多挑战，例如，宿主机、虚拟机管理器（VMM）或云服务商可能在未经授权的情况下访问用户数据，带来潜在的隐私泄露和安全风险，因此亟须引入更强的数据隔离机制以及可信的运行环境，来应对新一代云计算对数据安全的更高要求。
将机密计算技术应用于虚拟化领域，催生了机密虚拟化技术的诞生。机密虚拟化通过将机密计算构建的硬件级可信执行环境用于承载虚拟机，确保在即使多租户的云环境下，租户的数据和业务逻辑也能始终处于被机密计算保护的状态，有效消除传统虚拟化架构中的数据安全风险，提升云计算环境中的数据保护能力。这一技术的应用，使云服务提供商和企业租户能够在保障数据隐私的同时，灵活、高效地享受云计算带来的便捷和成本优势，推动了机密计算的普惠化和大众化。
顺应这种技术趋势，主流的处理器提供商，无论是x86、ARM还是RISC-V，都纷纷推出支持机密虚拟化的架构方案，以保障云端和本地计算环境中的数据机密性。同时，各大云服务提供商也在积极部署相关技术，国外的谷歌、Azure、AWS以及国内的阿里云、百度智能云、字节火山云等纷纷推出了涵盖机密虚拟机、机密容器以及异构机密实例在内的多种机密计算的云服务能力，旨在满足企业在互联网应用、机器学习、模型推理、云数据库等不同场景下对数据安全和隐私保护的迫切需求。
2024年，伴随着终端场景中大模型应用的兴起以及从业者对数据安全和隐私保护的日益关注，苹果公司推出了私人云计算（Private Cloud Computing，PCC），首次将机密计算的需求透射到数以亿计的手机终端的消费者市场。今天，无论是“数据可用不可见”的理念，还是隐私计算与机密计算的技术实现，已不再是曲高和寡、养在深闺人未知的神奇魔法，而是逐渐成为普罗大众触手可及且能广泛应用的普惠能力，不仅为个人和企业提供了更加安全和高效的计算体验，也为数据隐私和安全带来前所未有的保障。
中国作为全球最大的互联网市场，数字化转型的步伐不断加快，推动数据要素市场做大做强已成为经济转型成功的关键所在。在这一过程中，提升数据安全和隐私保护能力，正逐渐成为构建数据要素市场，保障数字经济健康发展的核心要素。机密虚拟化技术作为应对数据保护需求的重要手段，在中国市场正展现出巨大的潜力和广阔的应用前景。
自2021年起，本书作者所在的团队便与国内多家头部互联网企业紧密合作，共同推进机密虚拟化技术的产品研发、完善及业务落地。在实践过程中，我们深刻感受到中国这一庞大市场孕育出的多样、复杂且具有前瞻性的需求，不仅为机密计算技术的发展提供了肥沃的土壤，也持续推动着其技术理论与应用模式的持续创新。为更好应对这一趋势，一本能够系统阐述机密虚拟化的技术背景、实现原理及业务应用的专业著作显得尤为重要。
为什么撰写本书
机密虚拟化作为保障云上数据安全的关键技术，近年来受到广泛关注。然而，其体系庞杂，涉及多个技术领域，学习门槛较高。作为长期深耕于该领域的从业者，我们在实践中积累了丰富的经验和技术心得。在研究和实践过程中，我们也深感市面上缺乏一本系统、深入且注重实操的关于该领域的技术参考书。
为此，我们希望从实践视角出发，整理长期积累的认知与应用经验，撰写本书，全面解析机密虚拟化的核心原理与应用实践，阐明其在数据安全与隐私保护中的关键价值，帮助读者更好地理解、掌握并应用这项重要技术。为确保内容的系统性与前沿性，我们查阅了大量国内外资料，跟踪行业动态，最终完成了本书的写作。限于时间与能力，书中难免存在不足，恳请读者批评指正。
本书特色鲜明，知识体系清晰，从云时代对数据安全与隐私保护的基础需求出发，层层递进，深入剖析机密虚拟化技术的演进路径、架构原理、工程实现及应用案例，娓娓道来，力求为读者勾勒出一幅理论与实践并行的技术全景图。一方面，注重理论与工程实践结合，考虑到机密虚拟化的技术复杂性，选取英特尔可信域扩展（TDX）为核心案例，系统剖析其微架构设计、指令机制、高级特性及软件生态，帮助读者真正做到“知其然，更知其所以然”。另一方面，紧贴行业需求，结合联邦学习、大模型、异构计算、云数据库、区块链等典型应用，全面展示机密计算，尤其是机密虚拟化技术在云计算与大模型时代的应用模式与行业价值。
无论是对于初入门的技术爱好者，还是对于有意深入探索的专业人士，本书都将是一本极具参考价值的实用指南。
如何阅读本书
这是一本关于机密计算技术的专业书籍，内容丰富且具有深度，读者该如何高效利用这本书呢？
在阅读本书前，建议读者具备数据安全、云计算、计算机技术等相关基础知识，对常见的加密技术、虚拟化概念有一定了解，这样更有助于理解书中内容。
本书采用由浅入深的结构，按“基础概念 → 架构实现 → 应用案例 → 未来展望”的顺序编排，适合初学者循序渐进地学习。为充分吸收书中知识，建议按照以下三个阶段进行阅读。
第一阶段：整体通读。
先通读全书，重点理解数据安全、机密计算与机密虚拟化的基本概念与总体框架。对于技术细节和复杂实现，可以先略读，为后续深入学习打下基础。
第二阶段：深入剖析。
以Intel TDX技术为主线，深入理解机密虚拟化的微架构、指令系统和安全机制，并结合AMD、ARM等技术的实现，分析不同技术路径的差异与共性，形成系统认知。
第三阶段：针对性研读。
结合自身兴趣或项目需求，有选择地阅读相关章节。例如从事大模型应用的读者可重点关注对应章节的可信大模型方案以及异构机密计算的内容；关注云数据库安全的读者可深入研究机密数据库技术与方案。
若读者已具备机密计算领域的实践经验，则可按需跳读，直接查阅感兴趣的技术细节或案例内容。值得强调的是，学习机密计算不仅需要阅读，更需要实践。建议读者在研读本书的同时，结合实验环境、开源项目或实际系统，动手构建和验证相关技术，持续探索，不断深化理解与应用能力。

本书得以顺利付梓，离不开多方的鼎力支持。在此，谨向所有为本书提供指导、审阅以及帮助的人士致以诚挚谢意。特别要感谢参与本书审稿的国内外专家（名单按笔画和字母顺序排列）：于国瑞、刘双、刘煜堃、龙勤、田洪亮、闫守孟、张尧、张佳、张锋巍、汪晟、李志明、郑然、周翔、周爱辉、杨继国、胡潇、郭伟、Mohan J. Kumar 、Mona Vij、Simon Johnson等，感谢他们在百忙之中不辞辛劳、拨冗审阅书稿，提出了许多宝贵而深刻的意见，或慨然作序，或惠赐寄语，或细致批阅。专家们的专业视角与真知灼见，对提升本书内容质量与专业水平起到了关键作用。同时，特别感谢电子工业出版社的出版团队及本书责任编辑宋亚东先生，感谢他们严谨细致的工作与全程支持。还要对在成书过程中给予我们关心和帮助的各位同仁与朋友，致以诚挚谢意。
在本节完成之际，2025年火山引擎春季FORCE原动力大会正在北京召开。会上，字节跳动火山引擎正式发布了Jeddak AICC，以及联想依托该技术推出的面向大模型应用的PC终端领域首个“可信个人云”。还记得两年前，阿里云发布国内首个机密虚拟化的计算实例产品时，产品负责人曾提出，“机密计算应该是普惠的，是通用化、平民化的”，当时坐在台下的我深以为然。短短数年，国内的机密计算市场经历了快速而深刻的变革：从IaaS到PaaS，到SaaS，再到MaaS，部署形态和应用场景日益丰富；从数据库、科学计算，到大模型、生成式AI，再到如今终端智能体场景下的“可信个人云”，无不体现出数据安全与隐私保护已成为这个时代的基本诉求，也彰显了中国市场对创新技术的高度包容与快速响应。
在这一路的探索与实践中，我有一个愈发深刻的体会：我们始终被客户、被需求“推着”走。这种推力不仅源于技术自身的演进，更来自市场和需求的内在驱动，带来前所未有的紧迫感与现实感。正是这股力量，让技术创新不再抽象或枯燥，而是变得真实、具体、有温度。也正因此，我们更应以坚定的技术信念和务实的行动，持续前行，全力以赴。
宋川 朱运阁
2025年6月

查看更多 > 图书评论

暂无评论

发表图书评论

评论标题：
评论内容：
验证码：: 看不清楚
点击刷新
: 您还没有登录，请登录后再评论。
点击“登录”

购买过本书的顾客还买过

{{data.bookName}}

￥{{data.price}}