查看購物車
  • English
  • Japanese
  • Korean
封面
市場調查報告書
商品編碼
2064029

人工智慧定義車輛（AIDV）OEM廠商部署策略研究報告（2026年）

AI-Defined Vehicle (AIDV) OEMs' Deployment Strategies Research Report, 2026
出版日期: | 出版商: ResearchInChina | 英文 640 Pages | 商品交期: 最快1-2個工作天內

價格
簡介目錄

AIDV調查：22家汽車製造商的採用策略

由ResearchInChina發布的「人工智慧定義汽車（AIDV）OEM廠商部署策略研究報告（2026）」分析了包括理想汽車、蔚來汽車、小鵬汽車、小米汽車和吉利汽車在內的22家企業的AI部署策略。分析內容涵蓋了AI數據、雲端運算能力、車載運算能力、自主研發晶片、AI操作系統和AI平台模型在自動駕駛和智慧駕駛座等領域的應用，以及車載AI晶片的配置策略和規劃。

1. 新興 OEM：全面部署，從車輛擴展到 EAI。

目前，中國新能源汽車智慧化正進入下一階段，人工智慧技術已成為汽車製造商之間競爭的關鍵因素。

在2026年6月初舉行的高通中國汽車高峰會上，蔚來汽車創辦人李斌表示，如今的汽車製造商必須轉型為人工智慧公司，如今的智慧駕駛座必須轉型為人工智慧駕駛座。蔚來將智慧駕駛座的演進分為三個階段：「功能駕駛座」、「智慧型駕駛座」和「認知駕駛座」。

2026年1月，李翔在一次內部員工會議上提出了三點。第一，2026年將是志在成為人工智慧產業領導者的最後黃金時期。第二，L4級自動駕駛最遲將於2028年實用化。第三，全球能夠同時部署基礎模型、晶片、作業系統和企業應用人工智慧（EAI）的公司將不到三家，理想汽車將力爭成為其中之一。

2025 年 11 月，小鵬汽車在人工智慧日上宣布，將徹底升級其定位，成為「物理人工智慧領域的行動探索者和全球企業人工智慧公司」。

毫無疑問，人工智慧代理將成為2026年汽車產業的熱門詞彙之一。許多業內專家開始呼籲對圍繞人工智慧的過高期望保持理性。然而，這並不會阻止原始設備製造商（OEM）成為人工智慧驅動車輛（AI-V）領域最前沿的實踐者。與老牌OEM和國際品牌相比，新興OEM正在人工智慧運算能力、人工智慧晶片、人工智慧平台模型和企業人工智慧（EAI）等領域進行全面部署。

2. 傳統獨立品牌：關鍵突破集中在車輛智慧方面

吉利、長城汽車、長安、比亞迪和奇瑞等傳統獨立品牌正利用其對汽車製造的深刻理解，悄悄踏上「汽車智慧化」之路。它們正將人工智慧從一個單純的概念轉化為具體的科技突破，將一個宏偉的願景轉化為使用者可以體驗到的實際價值。

比亞迪發布了名為「玄機」的車載智慧架構。該架構利用規模經濟提供綜合技術，透過「車載智慧」架構促進電氣化和智慧化的深度融合，並專注於成本最佳化和全球適應性。

吉利選擇了「人工智慧集中化」的道路，專注於建立統一的「通用汽車大腦」。這將使他們能夠將人工智慧能力深度整合到駕駛座、自動駕駛和動力傳動系統等領域，從而實現跨領域協作和智慧演進。

傳統汽車製造商擁有數十年的汽車製造經驗，對車輛機械結構、動力系統、底盤調校和材料技術等領域的AI基礎知識有著深刻的理解，並具備成熟的工程實施能力。他們正利用AI建構車輛智慧架構，將自動駕駛、駕駛座、動力傳動系統系統、動力傳動系統和互聯等各個領域連接起來，實現跨領域整合。

3. 一些國際OEM廠商：對海外市場保持謹慎態度，深化與中國市場的合作。

國際汽車品牌正在採用明確的雙管齊下策略：一方面，針對中國市場和全球市場進行並行開發；另一方面，採取雙管齊下的方法，將與當地供應鏈的合作與核心技術的自主研發相結合；同時，透過推廣新能源汽車和傳統燃料汽車來實現兩輪驅動的成長。

2026年3月，BMW集團執行長宣布暫時降低全球市場L3級自動駕駛專案的研發優先順序。同時，在中國市場，寶馬集團發布了「360度全鏈AI戰略」，並與中國科技公司摩登科技合作，共同開發面向中國市場的全場景智慧駕駛系統。此外，基於阿里巴巴的大規模AI語言模型和DeepSeek的深度思考能力，「BMW智慧個人助理」升級為「AI個人助理」。

2026年1月，海外媒體報道稱，梅賽德斯-奔馳已暫時擱置在全球市場推廣L3級自動駕駛系統的計劃，轉而專注於L2+級自動駕駛系統。在中國市場，賓士自2017年起便與Momenta展開合作，並持續加大對該計畫的投資。兩家公司聯合開發的先進自動駕駛系統已應用於CLA和GLC SUV電池式電動車以及下一代S-Class轎車。

大眾汽車集團已製定「立足中國，服務中國」的策略。從2026年下半年開始，基於CEA（中國電子架構）的新車型將逐步搭載全球人工智慧（AI）智慧體。福斯與小鵬汽車共同開發的首款車型UNYX 08，搭載了兩顆小鵬圖靈AI晶片（總合力達1500 TOPS）。未來，兩家公司CEA架構與圖靈晶片的合作將應用於更多大眾品牌車型。

特斯拉在國際汽車品牌中獨樹一格。其人工智慧應用策略與新興的中國汽車製造商類似，都秉持著「人工智慧讓一切皆有可能」的理念。特斯拉將汽車視為「帶輪子的機器人」，將人形機器人視為「沒有輪子的汽車」。透過自主研發晶片和人工智慧平台模型，特斯拉實現了汽車和人機互動（EAI）共用同一套人工智慧技術堆疊。

目錄

定義

第1章：人工智慧定義車輛概述

  • 人工智慧定義車輛概述
  • 人工智慧定義的車輛與軟體定義的車輛
  • 人工智慧定義車輛的三個關鍵要素
  • 人工智慧正在改變汽車產業。
  • 人工智慧技術賦能整個OEM廠商的整個供應鏈（研發、生產、銷售、服務和供應鏈管理）。
  • 人工智慧技術在汽車生產的應用
  • 人工智慧技術在汽車生產中的應用：來自不同汽車製造商的應用案例總結。
  • 人工智慧技術在銷售和服務領域的應用
  • 人工智慧技術在銷售和服務領域的應用：OEM應用案例概述
  • AITV 數據統計
  • 人工智慧驅動的自動駕駛：2025 年的部署規模和滲透率
  • 人工智慧驅動的自動駕駛：2025 年預計普及率（按價格區間分類）
  • 人工智慧驅動的自動駕駛：2025 年預計滲透率（按車輛類型分類）
  • 人工智慧驅動的自動駕駛：2025 年預計滲透率（按新能源汽車類型分類）
  • 人工智慧驅動的自動駕駛：預計到 2025 年的採用率（按原始設備製造商分類）
  • 人工智慧驅動的自動駕駛：2025 年的普及率（按品牌分類）
  • 人工智慧驅動的自動駕駛：2025 年預計滲透率（按車輛類型分類）
  • 人工智慧驅動的行動號碼可攜性：2025 年的安裝量和普及率
  • 人工智慧驅動的行動號碼可攜性（MNP）：2025 年預計普及率（按價格區間分類）
  • 人工智慧驅動的行動支付：2025 年預計滲透率（按車輛類型分類）
  • 人工智慧驅動的行動網路：2025 年預計滲透率（按新能源汽車類型分類）
  • 人工智慧驅動的行動支付：2025 年預計滲透率（按原始設備製造商、品牌和車輛類型分類）
  • 人工智慧停車場：2025 年的安裝數量和滲透率
  • 人工智慧停車系統：2025 年預計滲透率（按價格區間分類）
  • 人工智慧泊車系統：2025 年預計滲透率（按車輛類型分類）
  • 人工智慧輔助泊車系統：2025 年預計滲透率（按新能源汽車類型分類）
  • 人工智慧泊車系統：2025 年預計普及率（按原始設備製造商、品牌和車輛類型分類）
  • 人工智慧語音助理：2025 年的安裝量和普及率
  • 人工智慧語音助理：2025 年預計滲透率（按價格區間分類）
  • 2025年人工智慧語音助理普及率（依車輛類型分類）
  • 人工智慧語音助理：2025 年預計滲透率（按新能源汽車類型分類）
  • 人工智慧語音助理：2025 年預計滲透率（按 OEM 廠商分類）
  • 人工智慧語音助理：2025 年普及率（按品牌分列）
  • 人工智慧語音助理：2025 年普及率（按車輛類型分類）
  • 人工智慧汽車基礎架構模型：2025 年的部署規模與採用率
  • 人工智慧驅動的汽車平台模型：2025 年預計滲透率（按價格區間分類）
  • 人工智慧驅動的汽車平台模型：2025 年預計滲透率（按車型分類）
  • 人工智慧驅動的汽車基礎車型：2025 年預計滲透率（按新能源汽車類型分類）
  • 人工智慧驅動的汽車平台模型：2025 年預計滲透率（按 OEM 廠商分類）
  • 人工智慧驅動的汽車平台模型：2025 年的採用率（按品牌分類）
  • 人工智慧驅動的汽車平台模型：2025 年預計滲透率（按車型分類）
  • 人工智慧控制車輛面臨的挑戰
  • 人工智慧汽車面臨的挑戰：技術難題及解決方案
  • 人工智慧控制車輛面臨的挑戰：社會倫理
  • 人工智慧汽車面臨的挑戰：業界標準
  • 人工智慧賦能車輛面臨的挑戰：法規
  • AI 定義汽車面臨的挑戰：引入汽車 AI 平台模型面臨的挑戰。
  • 各OEM公司的AI專利
  • 吉利——多智慧體協同車輛控制解決方案
  • 吉利：基於多模態記憶支援的自動駕駛方法
  • 吉利：汽車模型部署框架與模型存取方法
  • 廣汽集團：基於代理商的車輛運算任務調度解決方案
  • 廣汽集團：「駕駛座與駕駛座椅整合」的多智慧體協同控制解決方案
  • 長城汽車：基於代理的現役車輛控制解決方案
  • Chery：基於基礎模型意圖識別的多智慧體協作解決方案。
  • 長安：針對多輛車的協同決策解決方案
  • 長安：自動駕駛系統及其自主學習方法
  • 一汽集團：基於人工智慧代理的車輛故障分析解決方案
  • 一汽集團：汽車人工智慧與使用者的互動式開發方法
  • Leapmotor：一種基於基準模型和強化學習智慧體的知覺任務模型動態最佳化方法
  • Voyah：一種提高智慧駕駛座語音控制反應速度的解決方案。
  • 人工智慧控制車輛的發展趨勢
  • 趨勢 1：可解釋的大規模 AI 模型正在應用於汽車。
  • 趨勢二：汽車人工智慧應用泛化能力提升
  • 趨勢三：將人工智慧模組引入車輛將加速車輛智慧化進程。
  • 趨勢四：人工智慧與車輛作業系統的融合
  • 趨勢五：降低終端運算功耗的新技術
  • 趨勢 6：全球模型正在從輔助工具升級為核心基礎設施。
  • 趨勢7：VLA和全球模型向物理人工智慧趨同

第2章：人工智慧的基本策略與結構：資料 + 運算能力

  • 人工智慧定義的車輛：數據策略
  • 人工智慧在車輛資料擷取、傳輸和儲存的應用
  • 人工智慧在車輛資料處理、標註和訓練的應用
  • 各OEM廠商雲端平台整合概覽
  • OEM廠商雲端原生應用程式概覽
  • 人工智慧資料中心分佈情況（1）
  • 人工智慧資料中心分佈（2）
  • 公共雲端資料中心佈局
  • 供應商人工智慧數據利用案例匯總
  • 人工智慧定義的車輛：運算能力策略
  • 人工智慧技術應用和解決方案對雲端運算能力的要求
  • OEM廠商將如何建構人工智慧所需的雲端運算能力？
  • 國內OEM廠商雲端運算平台概覽
  • 主要汽車製造商代表性車型的車載運算能力配置（1）：理想汽車
  • 來自不同汽車製造商的代表性車型中的汽車計算能力配置（2）：小鵬汽車
  • 主要汽車製造商代表性車型的車載運算能力配置（3）：蔚來汽車
  • 代表性OEM車型車載運算能力配置（4）：小米汽車、躍躍欲試
  • 主要汽車製造商（OEM）代表性車型的車載運算能力配置（5）：比亞迪
  • 主要汽車製造商（OEM）代表性車型車載運算能力配置（6）：長安、北汽
  • 主要汽車製造商（OEM）代表性車型的車載運算能力配置（7）：長城汽車、上汽集團
  • 主要汽車製造商（OEM）代表性車型的車載運算能力配置（8）：奇瑞
  • 主要汽車製造商（OEM）代表性車型的車載運算能力配置（9）：吉利
  • 主要汽車製造商（OEM）代表性車型的車載運算能力配置（10）：東風汽車、廣汽集團
  • 主要汽車製造商（11家公司）代表性車型的車載運算能力配置：特斯拉、奧迪
  • 主要汽車製造商（12家公司）代表性車型的車載運算能力配置：豐田、日產
  • 自動駕駛車輛運算能力代幣收費模式
  • 人工智慧技術應用和解決方案中對車輛運算能力的要求
  • 汽車製造商將如何建構人工智慧所需的車輛運算能力？

第3章：人工智慧模型的策略與結構

  • 汽車產業基本模型應用概述
  • 人工智慧平台模型的定義和特徵
  • 人工智慧模型分類及其在汽車領域的應用
  • 駕駛座整合中央運算架構為實現人工智慧定義的車輛提供了有利環境（1）
  • 駕駛座整合中央運算架構為實現人工智慧定義的車輛提供了有利環境（2）
  • 車載設備功能呼叫模式（基於人工智慧平台模型）
  • OEM廠商中基於人工智慧的模型應用案例總結
  • 供應商人工智慧模型應用總結
  • 中國主流人工智慧模型概述
  • 人工智慧表格在汽車作業系統中的應用
  • 人工智慧模型對車輛作業系統的影響
  • AI Foundation 模型可用於產生 AUTOSAR 測試。
  • 人工智慧表格在車輛作業系統中的應用
  • AIOS架構：核心模組的關鍵元件和功能
  • AIOS架構：核心模組的關鍵元件和功能
  • AIOS架構：AI平台模型與任務流程的部署
  • AIOS架構：不同AI執行時間的比較
  • 人工智慧表格在自動駕駛中的應用
  • 人工智慧系統在智慧駕駛的應用
  • 基於人工智慧的生成式模擬技術增強了模擬系統的駕駛能力。
  • 人工智慧模型在智慧駕駛感知的應用
  • 人工智慧模型在智慧駕駛決策中的應用
  • 人工智慧模型在智慧駕駛的應用趨勢
  • 汽車製造商在自動駕駛領域實施人工智慧的策略
  • 人工智慧模型在智慧駕駛座的應用
  • AI模型在智慧駕駛座的應用：AI駕駛座與軟體定義駕駛座。
  • 人工智慧模型在智慧駕駛座中的應用場景
  • 智慧駕駛座中實施的人工智慧平台模型是標準化的。
  • 智慧駕駛座中基於人工智慧的模型的部署，其特點是邊緣優先技術和邊緣雲端的融合。
  • 在HUD中應用基於人工智慧的表格
  • 人工智慧模型在智慧駕駛座語音對話中的應用
  • 基於人工智慧模型的智慧駕駛座手勢姿態辨識應用
  • 人工智慧模型在DMS的應用
  • 人工智慧模型在智慧駕駛座個人化服務的應用
  • 人工智慧模型在智慧駕駛座的應用趨勢
  • OEM廠商在智慧駕駛座中實施AI的策略
  • 人工智慧代理在智慧車輛的應用
  • 汽車製造商在車輛中部署人工智慧代理的概述

第5章：通用人工智慧的層次：主要應用挑戰

  • 駕駛座L3代理迭代的四個階段
  • 駕駛座代理是創造使用者價值的基礎。
  • 駕駛艙代理：設備雲端架構的普及
  • 駕駛座代理：對於某些任務，設備端人工智慧較為合適。
  • 駕駛座代理：基本款進一步推進了駕駛座與駕駛座椅的整合。
  • 駕駛座代理：原始設備製造商的多種開發模式
  • 駕駛座代理：多模態/全模式
  • 駕駛座代理：典型的多模態架構
  • 具有湧現能力的智慧體：互動式智慧體的訓練方法
  • 擁有全新能力的智慧體：從類比到真實世界

第4章：人工智慧晶片策略與佈局

  • 部署邊緣人工智慧的晶片策略
  • 部署邊緣人工智慧：高效能、低功耗的記憶體內運算晶片
  • 邊緣人工智慧的應用：降低基於蒸餾技術的車輛的運算能力需求。
  • 邊緣人工智慧的實施：人工智慧智慧駕駛座的應用場景
  • 部署邊緣人工智慧：自動駕駛人工智慧的運算能力分配機制
  • 晶片供應商正透過多種管道實現深度邊緣人工智慧最佳化。
  • 邊緣基礎模型推動記憶體晶片升級，LPDDR6 成為下一代技術的重點。
  • 邊緣人工智慧部署：主流晶片和人工智慧運算能力
  • 邊緣人工智慧部署面臨的挑戰（1）：運算能力
  • 邊緣人工智慧部署面臨的挑戰（2）：存儲
  • 邊緣人工智慧部署面臨的挑戰（3）：大規模生產和部署中的DRAM儲存頻寬
  • 邊緣人工智慧部署的創新策略（1）：車載人工智慧代理
  • 邊緣人工智慧部署的創新策略（2）：AI Box
  • 邊緣人工智慧部署的創新策略（3）：車載人工智慧盒的部署模式
  • 邊緣人工智慧部署的創新策略（4）：幾款OEM人工智慧盒子產品的比較
  • OEM廠商邊緣AI部署策略
  • 支援智慧駕駛人工智慧的代表性晶片產品
  • 高性能晶片產品及其在汽車應用中的整合概述
  • 支援人工智慧的領先高效能晶片產品比較（1）：NVIDIA Thor-X
  • 支援人工智慧的領先高效能晶片產品比較（2）：NVIDIA Thor-U
  • 支援人工智慧的領先高效能晶片產品比較（3）：高通8797
  • 支援人工智慧的領先高效能晶片產品對比（4）：Horizo​​​​n J6P
  • 支援人工智慧的領先高性能晶片產品對比（5）：小鵬汽車的「圖靈」人工智慧晶片
  • 支援人工智慧的領先高效能晶片產品比較（6）：蔚來神機NX9031
  • 支援人工智慧的領先高性能晶片產品對比（7）：犀牛光智R1
  • 支援人工智慧的領先高效能晶片產品比較（8）：SiEngine「星辰1」（AD1000）
  • 支援人工智慧的領先高效能晶片產品比較（9）：安霸 CV3-AD685
  • 支援人工智慧的領先高效能晶片產品比較（10）：Nvidia Orin-Y
  • 支援人工智慧的領先高效能晶片產品比較（11）：NVIDIA OrinX
  • 支援人工智慧的領先高效能晶片產品比較（12）：特斯拉 HW 4.0 Gen 2 FSD
  • 支援人工智慧的領先高效能晶片產品對比（13）：瑞薩 R-Car X5H
  • 支援人工智慧的領先高效能晶片產品比較（14）：華為昇騰610
  • 支援人工智慧的領先高效能晶片產品比較（15）：黑芝麻A2000
  • 支援智慧駕駛座人工智慧的代表性晶片產品
  • 支援智慧駕駛座人工智慧的晶片產品概述
  • 支援智慧駕駛座AI的晶片產品比較（1）：SemiDrive X10、三星V920、聯發科MT8676（1）
  • 支援智慧駕駛座AI的晶片產品比較（1）：SemiDrive X10、三星V920、聯發科MT8676（2）
  • 支援智慧駕駛座AI的晶片產品比較（2）：高通8397、英特爾Panther Lake汽車版、高通SA8775P（1）
  • 支援智慧駕駛座AI的晶片產品比較（2）：高通8397、英特爾Panther Lake汽車版、高通SA8775P（2）
  • 支援智慧駕駛座AI的晶片產品對比（3）：瑞薩R-Car X5H、NVIDIA Thor、高通SA8795P（1）
  • 支援智慧駕駛座AI的晶片產品對比（3）：瑞薩R-Car X5H、NVIDIA Thor、高通SA8795P（2）
  • 支援智慧駕駛座AI的晶片產品比較（4）：聯發科C-X1/CT-Y1，高通8797（1）
  • 支援智慧駕駛座AI的晶片產品比較（4）：聯發科C-X1/CT-Y1，高通8797（2）
  • 支援智慧駕駛座AI的晶片產品比較（5）：聯發科天璣P1 Ultra/S1 Ultra、CT-X1（1）
  • 支援智慧駕駛座AI的晶片產品比較（5）：聯發科天璣P1 Ultra/S1 Ultra、CT-X1（2）
  • 智慧汽車人工智慧晶片成本分析
  • 智慧汽車AI晶片成本構成（1）
  • 智慧汽車人工智慧晶片成本細分（2）
  • 智慧汽車人工智慧晶片成本構成（3）
  • 智慧型汽車AI晶片成本構成（4）：流片成本
  • 智慧汽車用人工智慧晶片的預計運輸價格
  • 汽車製造商的晶片配置策略和計劃
  • 汽車製造商的晶片配置策略與計畫（1）：蔚來、小鵬、躍遷
  • 汽車製造商的晶片配置策略與方案（2）：理想汽車、廣汽集團
  • 汽車製造商的晶片配置策略與計畫（3）：一汽、上汽
  • 汽車製造商晶片配置策略及方案（4）：北汽集團、長安汽車
  • 汽車製造商的晶片配置策略與方案（5）：長城汽車、東風汽車
  • 汽車製造商的晶片配置策略與計畫（6）：吉利、小米汽車
  • 汽車製造商的晶片配置策略與計畫（7）：奇瑞、特斯拉
  • 汽車製造商的晶片配置策略和計劃（8）：比亞迪、寶馬
  • 汽車製造商的晶片配置策略和計劃（9）：大眾、奧迪、梅賽德斯-奔馳

第5章：OEM廠商在人工智慧定義型汽車領域的進展與發展

  • Li Auto
  • NIO
  • Xpeng
  • Xiaomi
  • Geely
  • BYD
  • AITO
  • Changan Automobile
  • BAIC
簡介目錄
Product Code: ZXF017

AIDV Research: Deployment Strategies of 22 OEMs

The AI-Defined Vehicle (AIDV) OEMs' Deployment Strategies Research Report, 2026, released by ResearchInChina, analyzes the AI deployment strategies of 22 OEMs such as Li Auto, NIO, XPeng, Xiaomi Auto, and Geely, involving the application of AI data, cloud computing power, automotive computing power, self-developed chips, AI operating systems and AI foundation models in intelligent driving, intelligent cockpits, and other fields, as well as automotive AI chip configuration strategies and planning.

1. Emerging OEMs: Comprehensive deployment expands from vehicles to EAI

Currently, the intelligence of China's new energy vehicles is entering the next stage, and AI technology has become a core variable in the competition among OEMs.

At the Qualcomm China Automotive Summit in early June 2026, NIO founder Li Bin clearly stated that today's automotive companies must become AI companies, and today's intelligent cockpits must turn into AI cockpits. NIO divides the intelligent cockpit evolution into three stages: "functional cockpits", "intelligent cockpits" and "cognitive cockpits".

In January 2026, Li Xiang made three assertions at an internal staff meeting: First, 2026 marks the last window of opportunity for enterprises aiming to become top players in the AI industry; second, L4 intelligent driving will definitely be applied as late as 2028; third, there will be no more than three companies in the world that can deploy foundation models, chips, operating systems, and EAI at the same time, and Li Auto will strive to become one of them.

In November 2025, XPeng announced at the AI Day that it would comprehensively upgrade its positioning to "a mobility explorer in the physical AI world and a global EAI company".

There is no doubt that AI agents, one of the buzzwords in the automotive industry in 2026. Many industry experts have started to call for cooling down the AI hype. But this does not affect the emergence of OEMs as the most radical practitioners of AIDVs. Compared with traditional OEMs and international brands, emerging OEMs have made a comprehensive layout in the fields of AI computing power, AI chips, AI foundation models, EAI and other fields.

2. Traditional Independent Brands: Key Breakthroughs Focus on Vehicle Intelligence

Traditional independent brands represented by Geely, Great Wall Motor, Changan, BYD, and Chery are relying on their deep understanding of vehicle manufacturing to tacitly embark on a "vehicle intelligence" path. They have transformed AI from a concept into a concrete technological breakthrough, and from a grand vision into perceived user value.

BYD has released a vehicle intelligent architecture called "Xuanji", which relies on scale effects to offer inclusive technology. It uses the "vehicle intelligent" architecture to promote the deep integration of electrification and intelligence, emphasizing cost optimization and global adaptability.

Geely has chosen the "AI centralization" path and is committed to building a unified "vehicle universal cerebrum" to deeply penetrate AI capabilities into cockpit, intelligent driving, powertrain and other fields to achieve cross-domain collaboration and intelligent evolution.

With decades of expertise in vehicle manufacturing, traditional OEMs have a deep understanding of vehicle mechanical structure, powertrain system, chassis tuning, material technology and other AI carriers, and mature engineering implementation capabilities. They use AI to link the intelligent driving domain, cockpit domain, powertrain domain, chassis domain and connectivity domain to achieve a vehicle intelligent architecture with cross-domain integration.

3. Some International OEMs: Stay Cautious in Overseas Markets and Deepen Cooperation in the Chinese Market

International auto brands have adopted distinct two-pronged strategies: dual-track parallel development for the Chinese market and global markets; dual-line progress through cooperation with local supply chains and independent R&D of core technologies; and two-wheel driven growth by promoting new energy vehicles and traditional fuel vehicles simultaneously.

In March 2026, the CEO of the BMW Group announced that it would temporarily lower the R&D priority of the L3 project in the global market; at the same time, in the Chinese market, the BMW Group released a "360-degree full-chain AI strategy" and joined hands with the Chinese technology company Momenta to jointly develop an all-scenario intelligent driving system for the Chinese market. Besides, based on Alibaba's AI large language model and DeepSeek's deep thinking capabilities, the BMW Intelligent Personal Assistant was upgraded to an "AI Personal Assistant";

In January 2026, overseas media reported that Mercedes-Benz had temporarily shelved its L3 intelligent driving system promotion plan in the global market and instead focused on the L2+ intelligent driving system. In the Chinese market, Mercedes-Benz has continued to invest in the project in cooperation with Momenta since 2017. The high-level intelligent driving system developed by the two parties has been implemented in the battery-electric CLA, GLC SUV and next-generation S-class sedan.

The Volkswagen Group adopts the "In China, for China" strategy. Starting from the second half of 2026. New vehicle models based on the CEA (China Electronic Architecture) will gradually be equipped with full-domain AI agents. The first model CO-developed by Volkswagen and XPeng, UNYX 08, is equipped with two XPeng Turing AI chips (totaling 1500 TOPS of computing power). Their cooperative CEA and Turing chips will be applied to more Volkswagen-branded models in the future.

Tesla is an outlier among international automotive brands. Its AI deployment strategy is more similar to that of emerging OEMs in China, exhibiting a distinct characteristic of "AI enables everything". Tesla sees cars as "wheeled robots" and humanoid robots as "wheelless cars", and develops its own chips and AI foundation models, making cars and EAI share the same AI technology stack.

Table of Contents

Definitions

1 Overview of AI-Defined Vehicles

  • 1.1 Overview of AI-Defined Vehicles
  • AI-Defined Vehicles vs. Software-Defined Vehicles (1)
  • AI-Defined Vehicles vs. Software-Defined Vehicles (2)
  • Three Key Elements of AI-Defined Vehicles (1)
  • Three Key Elements of AI-Defined Vehicles (2)
  • AI Is Reshaping the Automotive Industry Pattern
  • AI Technology Empowers OEMs Across the Entire Chain: R&D, Production, Sales, Service, and Supply Chain Management
  • Application of AI Technology in Vehicle Production
  • Application of AI Technology in Vehicle Production: Summary of OEMs' Applications (1)
  • Application of AI Technology in Vehicle Production: Summary of OEMs' Applications (2)
  • Application of AI Technology in Sales and Service
  • Application of AI Technology in Sales and Service: Summary of OEMs' Applications
  • 1.2 AIDV Data Statistics
  • AI-powered Autonomous Driving: Installation Volume and Penetration Rate in 2025
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by Price Range)
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by Vehicle Class)
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by New Energy Vehicle Type)
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by OEM)
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by Brand)
  • AI-powered Autonomous Driving: Penetration Rate in 2025 (by Vehicle Model)
  • AI-powered MNP: Installation Volume and Penetration Rate in 2025
  • AI-powered MNP: Penetration Rate in 2025 (by Price Range)
  • AI-powered MNP: Penetration Rate in 2025 (by Vehicle Class)
  • AI-powered MNP: Penetration Rate in 2025 (by New Energy Vehicle Type)
  • AI-powered MNP: Penetration rate in 2025 (by OEM, Brand and Vehicle Model)
  • AI-powered Parking: Installation Volume and Penetration Rate in 2025
  • AI-powered Parking: Penetration Rate in 2025 (by Price Range)
  • AI-powered Parking: Penetration Rate in 2025 (by Vehicle Class)
  • AI-powered Parking: Penetration Rate in 2025 (by New Energy Vehicle Type)
  • AI-powered Parking: Penetration rate in 2025 (by OEM, Brand and Vehicle Model)
  • AI-powered Voice Assistant: Installation Volume and Penetration Rate in 2025
  • AI-powered Voice Assistant: Penetration Rate in 2025 (by Price Range)
  • AI-powered Voice Assistant Penetration Rate in 2025 (by Vehicle Class)
  • AI-powered Voice Assistant: Penetration Rate in 2025 (by New Energy Vehicle Type)
  • AI-powered Voice Assistant: Penetration Rate in 2025 (by OEM)
  • AI-powered Voice Assistant: Penetration Rate in 2025 (by Brand)
  • AI-powered Voice Assistant: Penetration Rate in 2025 (by Vehicle Model)
  • AI Automotive Foundation Models: Installation Volume and Penetration Rate in 2025
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by Price Range)
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by Vehicle Class)
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by New Energy Vehicle Type)
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by OEM)
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by Brand)
  • AI Automotive Foundation Models: Penetration Rate in 2025 (by Vehicle Model)
  • 1.3 Challenges in AI-Defined Vehicles
  • Challenges in AI-Defined Vehicles: Technical Difficulties and Solutions (1)
  • Challenges in AI-Defined Vehicles: Technical Difficulties and Solutions (2)
  • Challenges in AI-Defined Vehicles: Social Ethics
  • Challenges in AI-Defined Vehicles: Industry Standards
  • Challenges in AI-Defined Vehicles: Laws and Regulations (1)
  • Challenges in AI-Defined Vehicles: Laws and Regulations (2)
  • Challenges in AI-Defined Vehicles: Challenges in Deploying Automotive AI Foundation Models
  • 1.4 AI Patents of OEMs
  • Geely - Multi-Agent Collaborative Vehicle Control Solution
  • Geely: An Autonomous Driving Method Based on Multi-Modal Memory Assistance
  • Geely: An Automotive Model Deployment Framework and Model Access Method
  • GAC Group: An Agent-Based Vehicle Computing Task Scheduling Solution
  • GAC Group: A Multi-Agent Collaborative Control Solution for "Cockpit-Driving Integration"
  • Great Wall Motor: An Agent-Based Active Service Vehicle Control Solution
  • Chery: A Multi-Agent Collaboration Solution Based on Foundation Model Intent Recognition
  • Changan Automobile: A Multi-Autonomous-Vehicle Collaborative Decision-Making Solution
  • Changan Automobile: An Autonomous Driving System and Its Autonomous Learning Method
  • FAW Group: A Vehicle Fault Analysis Solution Based on an AI Agent
  • FAW Group: A Two-Way Development Method for Automotive AI and Users
  • Leapmotor: A Dynamic Optimization Method for Perception Task Models Based on Benchmark Models and Reinforcement Learning Agents
  • Voyah: A Solution to Improve the Response Speed ??of Voice Control in Intelligent Cockpits
  • 1.5 Development Trends of AI-Defined Vehicles
  • Trend 1: Explainable AI Large Models Are Being Applied to Vehicles (1)
  • Trend 1: Explainable AI Large Models Are Being Applied to Vehicles (2)
  • Trend 1: Explainable AI Large Models Are Being Applied to Vehicles (3)
  • Trend 2: Automotive AI Applications Increase Generalization Capabilities
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (1)
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (2)
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (3)
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (4)
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (5)
  • Trend 3: Introduction of AI BOX in Vehicles Accelerates Vehicle Intelligence (6)
  • Trend 4: AI Integrates with Vehicle Operating Systems (1)
  • Trend 4: AI Integrates with Vehicle Operating Systems (2)
  • Trend 4: AI Integrates with Vehicle Operating Systems (3)
  • Trend 4: AI Integrates with Vehicle Operating Systems (4)
  • Trend 4: AI Integrates with Vehicle Operating Systems (5)
  • Trend 5: New Technologies Lower End-side Computing Power Requirements
  • Trend 6: World Models Upgrade from Auxiliary Tools to Core Foundation (1)
  • Trend 6: World Models Upgrade from Auxiliary Tools to Core Foundation (2)
  • Trend 7: VLA and World Models Converge toward Physical AI (1)
  • Trend 7: VLA and World Models Converge toward Physical AI (2)
  • Trend 7: VLA and World Models Converge toward Physical AI (3)
  • Trend 7: VLA and World Models Converge toward Physical AI (4)
  • Trend 7: VLA and World Models Converge toward Physical AI (5)

2 AI Basic Strategies and Layout: Data + Computing Power

  • 2.1 AI-Defined Vehicles: Data Strategies
  • AI Applications in Vehicle Data Collection, Transmission, and Storage
  • AI Applications in Vehicle Data Processing, Annotation, and Training
  • Summary of OEMs' Cloud Platform Cooperation (1)
  • Summary of OEMs' Cloud Platform Cooperation (2)
  • Summary of OEMs' Cloud Platform Cooperation (3)
  • Summary of Cloud Native Application by OEMs (1)
  • Summary of Cloud Native Application by OEMs (2)
  • Distribution of AI Data Centers (1)
  • Distribution of AI Data Centers (2)
  • Public Cloud Data Center Layout
  • Summary of AI Data Application by Suppliers (1)
  • Summary of AI Data Application by Suppliers (2)
  • Summary of AI Data Application by Suppliers (3)
  • Summary of AI Data Application by Suppliers (4)
  • 2.2 AI-Defined Vehicles: Computing Power Strategies
  • Requirements for Cloud Computing Power in AI Technology Applications and Solutions
  • How OEMs Build Cloud Computing Power Required by AI (1)
  • How OEMs Build Cloud Computing Power Required by AI (2)
  • Summary of Cloud Computing Power Platforms of Domestic OEMs (1)
  • Summary of Cloud Computing Power Platforms of Domestic OEMs (2)
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (1): Li Auto
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (2): XPeng
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (3): NIO
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (4): Xiaomi Auto, Leapmotor
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (5): BYD
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (6): Changan, BAIC
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (7): Great Wall Motor, SAIC
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (8): Chery
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (9): Geely
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (10): Dongfeng Motor, GAC
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (11): Tesla, Audi
  • Automotive Computing Power Configuration of Typical Vehicle Models of OEMs (12): Toyota, Nissan
  • Autonomous Vehicle Computing Power Token Billing Modes (1)
  • Autonomous Vehicle Computing Power Token Billing Modes (1)
  • Requirements for Vehicle Computing Power in AI Technology Applications and Solutions
  • How OEMs Build Vehicle Computing Power Required by AI

3 AI Model Strategies and Layout

  • 3.1 Overview of Foundation Model Applications in the Automotive Industry
  • Definition and Features of AI Foundation Models
  • Classification of AI Foundation Models and Their Applications in the Automotive Sector
  • Cockpit-Driving Integration Central Computing Architecture Provides A Favorable Environment for Implementation of AI-Defined Vehicles (1)
  • Cockpit-Driving Integration Central Computing Architecture Provides A Favorable Environment for Implementation of AI-Defined Vehicles (2)
  • Invocation Modes of Automotive Device Functions by AI Foundation Models (1)
  • Invocation Modes of Automotive Device Functions by AI Foundation Models (2)
  • Summary of AI Foundation Model Applications of OEMs
  • Summary of AI Foundation Model Applications of Suppliers
  • Summary of Mainstream AI Foundation Models in China
  • 3.2 Application of AI Foundation Models in Automotive Operating Systems
  • Impacts of AI Foundation Models on Vehicle Operating Systems
  • AI Foundation Models Can Be Used to Generate AUTOSAR Tests
  • Application of AI Foundation Models in Vehicle Operating Systems (1)
  • Application of AI Foundation Models in Vehicle Operating Systems (2)
  • Application of AI Foundation Models in Vehicle Operating Systems (3)
  • Application of AI Foundation Models in Vehicle Operating Systems (4)
  • Application of AI Foundation Models in Vehicle Operating Systems (5)
  • AIOS Architecture: Main Components and Functions of Kernel Module (1)
  • AIOS Architecture: Main Components and Functions of Kernel Module (2)
  • AIOS Architecture: Main Components and Functions of Kernel Module (3)
  • AIOS Architecture: Main Components and Functions of Kernel Module (4)
  • AIOS Architecture: Main Components and Functions of Kernel Module (5)
  • AIOS Architecture: Main Components and Functions of Kernel Module (6)
  • AIOS Architecture: AI Foundation Model Deployment and Task Flow
  • AIOS Architecture: Comparison between Different AI Runtimes
  • 3.3 Application of AI Foundation Models in Autonomous Driving
  • Application of AI Foundation Models in Intelligent Driving
  • Generative Simulation Technology for AI Foundation Models Enhances Capabilities of Driving Simulation Systems
  • Application of AI Foundation Models in Intelligent Driving Perception (1)
  • Application of AI Foundation Models in Intelligent Driving Perception (2)
  • Application of AI Foundation Models in Intelligent Driving Decision (1)
  • Application of AI Foundation Models in Intelligent Driving Decision (2)
  • Trends in Application of AI Foundation Models in Intelligent Driving (1)
  • Trends in Application of AI Foundation Models in Intelligent Driving (2)
  • Trends in Application of AI Foundation Models in Intelligent Driving (3)
  • AI Deployment Strategies for Autonomous Driving of OEMs (1)
  • AI Deployment Strategies for Autonomous Driving of OEMs (2)
  • AI Deployment Strategies for Autonomous Driving of OEMs (3)
  • AI Deployment Strategies for Autonomous Driving of OEMs (4)
  • AI Deployment Strategies for Autonomous Driving of OEMs (5)
  • AI Deployment Strategies for Autonomous Driving of OEMs (6)
  • AI Deployment Strategies for Autonomous Driving of OEMs (7)
  • AI Deployment Strategies for Autonomous Driving of OEMs (8)
  • AI Deployment Strategies for Autonomous Driving of OEMs (9)
  • AI Deployment Strategies for Autonomous Driving of OEMs (10)
  • AI Deployment Strategies for Autonomous Driving of OEMs (11)
  • AI Deployment Strategies for Autonomous Driving of OEMs (12)
  • AI Deployment Strategies for Autonomous Driving of OEMs (13)
  • AI Deployment Strategies for Autonomous Driving of OEMs (14)
  • AI Deployment Strategies for Autonomous Driving of OEMs (15)
  • AI Deployment Strategies for Autonomous Driving of OEMs (16)
  • AI Deployment Strategies for Autonomous Driving of OEMs (17)
  • 3.4 Application of AI Foundation Models in Smart Cockpits
  • Application of AI Foundation Models in Intelligent Cockpit: AI-Defined Cockpit vs. Software-Defined Cockpit
  • Application Scenarios of AI Foundation Models in Intelligent Cockpit
  • AI Foundation Models Deployed In Smart Cockpits Have Become Standard
  • Deployment of AI Foundation Models in Smart Cockpits Features Edge-First, Edge-Cloud Collaboration
  • Application of AI Foundation Models in HUD
  • Application of AI Foundation Models in Intelligent Cockpit Voice Interaction (1)
  • Application of AI Foundation Models in Intelligent Cockpit Voice Interaction (2)
  • Application of AI Foundation Models in Intelligent Cockpit Gesture Recognition
  • Application of AI Foundation Models in DMS (1)
  • Application of AI Foundation Models in DMS (2)
  • Application of AI Foundation Models in Intelligent Cockpit Personalized Services
  • Trends in Application of AI Foundation Models in Intelligent Cockpit (1)
  • Trends in Application of AI Foundation Models in Intelligent Cockpit (2)
  • Trends in Application of AI Foundation Models in Intelligent Cockpit (3)
  • AI Deployment Strategies for Smart Cockpits of OEMs (1)
  • AI Deployment Strategies for Smart Cockpits of OEMs (2)
  • AI Deployment Strategies for Smart Cockpits of OEMs (3)
  • AI Deployment Strategies for Smart Cockpits of OEMs (4)
  • AI Deployment Strategies for Smart Cockpits of OEMs (5)
  • AI Deployment Strategies for Smart Cockpits of OEMs (6)
  • AI Deployment Strategies for Smart Cockpits of OEMs (7)
  • AI Deployment Strategies for Smart Cockpits of OEMs (8)
  • AI Deployment Strategies for Smart Cockpits of OEMs (9)
  • AI Deployment Strategies for Smart Cockpits of OEMs (10)
  • AI Deployment Strategies for Smart Cockpits of OEMs (11)
  • AI Deployment Strategies for Smart Cockpits of OEMs (12)
  • 3.5 Application of AI Agent in Intelligent Vehicles
  • Summary of AI Agent Deployment in Vehicles by OEMs (1)
  • Summary of AI Agent Deployment in Vehicles by OEMs (2)
  • Summary of AI Agent Deployment in Vehicles by OEMs (3)

5 Levels of AGI: Main Application Issues

  • Four Stages of Cockpit L3 Agent Iteration
  • Cockpit Agent Is the Foundation for Generating User Value
  • Cockpit Agent: Popularization of Device-cloud Architecture
  • Cockpit Agent: On-device AI Is More Suitable for Specific Tasks
  • Cockpit Agent: Foundation Models Further Advances Cockpit-Driving Integration
  • Cockpit Agent: Different Development Modes of OEMs
  • Cockpit Agent: Multimodal / Omnimodal
  • Cockpit Agent: Typical Multimodal Architecture (1)
  • Cockpit Agent: Typical Multimodal Architecture (2)
  • Agents with Emergent Capabilities: Training Methods of Interactive Agents
  • Agents with Emergent Capabilities: Sim-to-real

4 AI Chip Strategies and Layout

  • 4.1 Chip Strategies for Edge AI Deployment
  • Edge AI Deployment: High-Compute, Low-Power Computing-in-Memory Chips
  • Edge AI Deployment: Distillation Can Lower Vehicle Computing Power Requirements
  • Edge AI Deployment: Application Scenarios of AI Smart Cockpits
  • Edge AI Deployment: Autonomous Driving AI Computing Power Allocation Mechanism
  • Chip Vendors Fulfill Deep Edge AI Optimization Through Multiple Paths
  • Edge Foundation Models Drive Memory Chip Upgrades, And LPDDR6 Becomes the Next-Generation Technology Focus
  • Edge AI Deployment: Mainstream Chips and AI Computing Power (1)
  • Edge AI Deployment: Mainstream Chips and AI Computing Power (2)
  • Challenges for Edge AI Deployment (1): Computing Power
  • Challenges for Edge AI Deployment (2): Storage (1)
  • Challenges for Edge AI Deployment (2): Storage (2)
  • Challenges for Edge AI Deployment (3): DRAM Storage Bandwidth in Mass Production and Deployment (1)
  • Challenges for Edge AI Deployment (3): DRAM Storage Bandwidth in Mass Production and Deployment (2)
  • Challenges for Edge AI Deployment (3): DRAM Storage Bandwidth in Mass Production and Deployment (3)
  • Challenges for Edge AI Deployment (3): DRAM Storage Bandwidth in Mass Production and Deployment (4)
  • Challenges for Edge AI Deployment (3): DRAM Storage Bandwidth in Mass Production and Deployment (5)
  • Innovative Strategies for Edge AI Deployment (1): AI Agents Deployed in Vehicles
  • Innovative Strategies for Edge AI Deployment (2): AI Box (1)
  • Innovative Strategies for Edge AI Deployment (2): AI Box (2)
  • Innovative Strategies for Edge AI Deployment (3): Deployment Modes of AI Box in Vehicles
  • Innovative Strategies for Edge AI Deployment (4): Comparison between Some OEM AI Box Products (1)
  • Innovative Strategies for Edge AI Deployment (4): Comparison of Some OEM AI Box Products (2)
  • Edge AI Deployment Strategies of OEMs (1): Foundation Models Are Compressed for Vehicle Deployment to Achieve Edge-Cloud Integration (1)
  • Edge AI Deployment Strategies of OEMs (1): Foundation Models Are Compressed for Vehicle Deployment to Achieve Edge-Cloud Integration(2)
  • Edge AI Deployment Strategies of OEMs (2): Cross-Domain Integration SoC Performance Iteration Accelerates (1)
  • Edge AI Deployment Strategies of OEMs (2): Cross-Domain Integration SoC Performance Iteration Accelerates (2)
  • 4.2 Typical Chip Products That Support Intelligent Driving AI
  • Summary of High-compute Chip Products That Support AI, and Automotive Application Cooperation (1)
  • Summary of High-compute Chip Products That Support AI, and Automotive Application Cooperation (2)
  • Summary of High-compute Chip Products That Support AI, and Automotive Application Cooperation (3)
  • Comparison of Typical High-compute Chip Products That Support AI (1): NVIDIA Thor-X
  • Comparison of Typical High-compute Chip Products That Support AI (2): NVIDIA Thor-U
  • Comparison of Typical High-compute Chip Products That Support AI (3): Qualcomm 8797
  • Comparison of Typical High-compute Chip Products That Support AI (4): Horizon J6P
  • Comparison of Typical High-compute Chip Products That Support AI (5): Xpeng "Turing" AI chip
  • Comparison of Typical High-compute Chip Products That Support AI (6): NIO Shenji NX9031
  • Comparison of Typical High-compute Chip Products That Support AI (7): Rhino Guangzhi R1
  • Comparison of Typical High-compute Chip Products That Support AI (8): SiEngine "Xingchen No.1" (AD1000)
  • Comparison of Typical High-compute Chip Products That Support AI(9): Ambarella CV3-AD685
  • Comparison of Typical High-compute Chip Products That Support AI (10): Nvidia Orin-Y
  • Comparison of Typical High-compute Chip Products That Support AI (11): NVIDIA Orin-X
  • Comparison of Typical High-compute Chip Products That Support AI (12): Tesla HW 4.0 Gen 2 FSD
  • Comparison of Typical High-compute Chip Products That Support AI (13): Renesas R-Car X5H
  • Comparison of Typical High-compute Chip Products That Support AI (14): Huawei Ascend 610
  • Comparison of Typical High-compute Chip Products That Support AI (15): Black Sesame A2000
  • 4.3 Typical Chip Products That Support Intelligent Cockpit AI
  • Summary of Chip Products That Support Intelligent Cockpit AI
  • Comparison of Chip Products That Support Intelligent Cockpit AI (1): SemiDrive X10, Samsung V920, MediaTek MT8676 (1)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (1): SemiDrive X10, Samsung V920, MediaTek MT8676 (2)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (2): Qualcomm 8397, Intel Panther Lake Automotive Edition, Qualcomm SA8775P (1)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (2): Qualcomm 8397, Intel Panther Lake Automotive Edition, Qualcomm SA8775P (2)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (3): Renesas R-Car X5H, NVIDIA Thor, Qualcomm SA8795P (1)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (3): Renesas R-Car X5H, NVIDIA Thor, Qualcomm SA8795P (2)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (4): MediaTek C-X1/CT-Y1, Qualcomm 8797 (1)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (4): MediaTek C-X1/CT-Y1, Qualcomm 8797 (2)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (5): MediaTek Dimensity P1 Ultra/S1 Ultra, CT-X1 (1)
  • Comparison of Chip Products That Support Intelligent Cockpit AI (5): MediaTek Dimensity P1 Ultra/S1 Ultra, CT-X1 (2)
  • 4.4 Analysis of Intelligent Vehicle AI Chip Costs
  • Composition of Intelligent Vehicle AI Chip Costs (1)
  • Composition of Intelligent Vehicle AI Chip Costs (2)
  • Composition of Intelligent Vehicle AI Chip Costs (3)
  • Composition of Intelligent Vehicle AI Chip Costs (4): Tape-out Cost
  • Intelligent Vehicle AI Chip Shipping Price Estimate (1)
  • Intelligent Vehicle AI Chip Shipping Price Estimate (2)
  • Intelligent Vehicle AI Chip Shipping Price Estimate (3)
  • Intelligent Vehicle AI Chip Shipping Price Estimate (4)
  • Intelligent Vehicle AI Chip Shipping Price Estimate (5)
  • 4.5 OEMs' Automotive Chip Configuration Strategies and Planning
  • OEMs' Automotive Chip Configuration Strategies and Planning (1): NIO, XPeng, Leapmotor
  • OEMs' Automotive Chip Configuration Strategies and Planning (2): Li Auto, GAC Group
  • OEMs' Automotive Chip Configuration Strategies and Planning (3): FAW, SAIC
  • OEMs' Automotive Chip Configuration Strategies and Planning (4): BAIC Group, Changan
  • OEMs' Automotive Chip Configuration Strategies and Planning (5): Great Wall Motor, Dongfeng Motor
  • OEMs' Automotive Chip Configuration Strategies and Planning (6): Geely, Xiaomi Auto
  • OEMs' Automotive Chip Configuration Strategies and Planning (7): Chery, Tesla
  • OEMs' Automotive Chip Configuration Strategies and Planning (8): BYD, BMW
  • OEMs' Automotive Chip Configuration Strategies and Planning (9): Volkswagen, Audi, Mercedes-Benz

5 OEMs' Progress and Layout in AI-Defined Vehicles

  • 5.1 Li Auto
  • AI Strategy (1)
  • AI Strategy (2)
  • AI Data Strategy
  • AI Computing Strategy (1): Cloud Computing
  • AI Computing Strategy (2): Edge Computing
  • AI Chip Strategy (1): Self-developed Chip M100
  • AI Chip Strategy (2): Chip Cooperation
  • AI Deployment Strategies for Vehicle Operating Systems (1): Li OS
  • AI Deployment Strategies for Vehicle Operating Systems (2): Halo OS (1)
  • AI Deployment Strategies for Vehicle Operating Systems (2): Halo OS (2)
  • AI Deployment in Intelligent Driving (1): AI Large Model Iteration
  • AI Deployment in Intelligent Driving (2): Five Major Technological Innovations of MindVLA-o1
  • AI Deployment in Intelligent Driving (2): 3D Vision Model of MindVLA-o1
  • AI Deployment in Intelligent Driving (2): Predictive Latent World Model of MindVLA-o1
  • AI Deployment in Intelligent Driving (2): Action Architecture of MindVLA-o1 Large Model
  • AI Deployment in Intelligent Driving (2): Iteration Mode of MindVLA-o1 Large Model
  • AI Deployment in Intelligent Driving (3): Architecture of MindVLA-U1 Large Model
  • AI Deployment in Intelligent Driving (4): Physical AI
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (2): Mind GPT Architecture and Technical Features
  • AI Deployment in Intelligent Cockpits (2): Multimodal Perceptive Interaction Design Empowered by MindGPT
  • AI Deployment in Intelligent Cockpits (3): Lixiang Tongxue Evolves into A Life Assistant Agent
  • AI Deployment in Intelligent Cockpits (4): Real-time Voice Dialogue Large Model MindGPT-4o-Audio
  • 5.2 NIO
  • AI Strategy
  • AI Data Strategy
  • AI Computing Strategy (1): Cloud Computing
  • AI Computing Strategy (2): Edge Computing
  • AI Chip Strategy (1): Self-developed Chip Shenji NX9031
  • AI Chip Strategy (2): Chip Cooperation
  • AI Deployment Strategies for Operating Systems: SkyOS (1)
  • AI Deployment Strategies for Operating Systems: SkyOS (2)
  • AI Deployment in Intelligent Driving(1): Iteration of AI Large Models
  • AI Deployment in Intelligent Driving (2): Iterative Development of AI Foundation Model Intelligent Driving System
  • AI Deployment in Intelligent Driving (3): NADArch 2.0 Architecture
  • AI Deployment in Intelligent Driving (4): NIO World Model (1)
  • AI Deployment in Intelligent Driving (4): NIO World Model (2)
  • AI Deployment in Intelligent Driving (4): NIO World Model (3)
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (2): Banyan 3
  • AI Deployment in Intelligent Cockpits (3): NOMI GP
  • AI Deployment in Intelligent Cockpits (4): AI Emotional Engine
  • AI Deployment in Intelligent Cockpits (5): A-VL Technology
  • 5.3 Xpeng
  • AI Strategy: From Smart Cars to Physical AI
  • AI Data Strategy
  • AI Computing Power Strategy (1): Cloud Computing Power
  • AI Computing Power Strategy (2): Edge Computing Power
  • AI Chip Strategy (1): Self-Developed "Turing" Chip
  • AI Chip Strategy (2): Performance Parameters of Turing Chip
  • AI Chip Strategy (3): Chip Collaboration
  • AI Chip Strategy (4): Turing Chip First Installed in Xpeng G7 in 2025
  • AI Deployment Strategies for Operating Systems (1): Tianji AIOS
  • AI Deployment Strategies for Operating Systems (2): XOS Development Plan
  • AI Deployment Strategies for Operating Systems (3): Tianji System Is Upgraded to AIOS
  • AI Deployment Strategies for Operating Systems (4): Functional Modules of AI OS
  • AI Deployment in Intelligent Driving (1): Iteration of AI Large Models
  • AI Deployment in Intelligent Driving (2): Xnet, Xplanner, XBrain
  • AI Deployment in Intelligent Driving (3): AI+XNGP
  • AI Deployment in Intelligent Driving (4): World Foundation Model "Cloud Model Factory"
  • AI Deployment in Intelligent Driving (5): Second Generation VLA (1)
  • AI Deployment in Intelligent Driving (5): Second Generation VLA (2)
  • AI Deployment in Intelligent Driving (6): Fast Drive VLA Model (1)
  • AI Deployment in Intelligent Driving (6): Fast Drive VLA Model (2)
  • AI Deployment in Intelligent Driving (7): World Model X-World (1)
  • AI Deployment in Intelligent Driving (7): World Model X-World (2)
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (2): Evolution of Intelligent Cockpit System
  • AI Deployment in Intelligent Cockpits (3): Core Functions of AI Cockpit
  • AI Deployment in Intelligent Cockpits (4): AI Cockpit-Driving Integration
  • AI Deployment in Intelligent Cockpits (5): AIOS 6.0
  • AI Deployment in Intelligent Cockpits (6): XPeng VLM
  • 5.4 Xiaomi
  • AI Strategy: Focus on Deploying Infrastructure Including AI Large Models
  • AI Data Strategy
  • AI Computing Power Strategy: Edge and Cloud Computing Power
  • AI Chip Strategy: Chip Cooperation
  • AI Deployment Strategies for Operating Systems (1): HyperOS (1)
  • AI Deployment Strategies for Operating Systems (1): HyperOS (2)
  • AI Deployment Strategies for Operating Systems (2): Accessing DeepSeek R1
  • AI Deployment Strategies for Operating Systems (3): HyperOS 2.0
  • AI Deployment in Intelligent Driving (1): Iteration of AI Large Models
  • AI Deployment in Intelligent Driving (2): Xiaomi Pilot
  • AI Deployment in Intelligent Driving (3): MiLM
  • AI Deployment in Intelligent Driving (4): Vision Language Model (VLM) (1)
  • AI Deployment in Intelligent Driving (4): Vision-Language Model (VLM) (2)
  • AI Deployment in Intelligent Driving (4): Vision-Language Model (VLM) (3)
  • AI Deployment in Intelligent Driving (5): Hyper Autonomous Driving Enhanced Edition
  • AI Deployment in Intelligent Driving (6): Genesis World Model (1)
  • AI Deployment in Intelligent Driving (6): Genesis World Model (2)
  • AI Deployment in Intelligent Driving (7): Xiaomi XLA
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (2): AI Voice
  • AI Deployment in Intelligent Cockpits (3): Connecting Alibaba Tongyi Large Model
  • AI Deployment in Intelligent Cockpits (4): HyperOS Smart Cabin
  • 5.5 Geely
  • AI Strategy: Full-Domain AI (1)
  • AI Strategy: Full-Domain AI (2)
  • AI Data Strategy
  • AI Computing Power Strategy (1): Cloud Computing Power
  • AI Computing Power Strategy (2): Edge Computing Power
  • AI Chip Strategy (1): Self-developed Chip Planning
  • AI Chip Strategy (2): Self-developed Chip Longying No.1 (SE1000)
  • AI Chip Strategy (3): Chip Cooperation
  • AI Deployment Strategies for Operating Systems (1): Flyme AI OS
  • AI Deployment Strategies for Operating Systems (2): ZEEKR AI OS
  • AI Deployment Strategies for Operating Systems (3): SOA Atomic Service
  • AI Deployment Strategies for Operating Systems (4): Upgrading AIOS Vehicle Intelligent Operating System
  • AI Deployment in Intelligent Driving (1): AI Foundation Model Iteration
  • AI Deployment in Intelligent Driving (2): E2E-based Haohan Intelligent Driving 2.0 Solution (1)
  • AI Deployment in Intelligent Driving (2): E2E-based Haohan Intelligent Driving 2.0 Solution (2)
  • AI Deployment in Intelligent Driving (3): Xingrui Large Model + DeepSeek-R1
  • AI Deployment in Intelligent Driving (4): VLA+World Generation Model+AI Drive Large Model
  • AI Deployment in Intelligent Driving (5): World Action Model
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (3): AI Agents
  • AI Deployment in Intelligent Cockpits (4): Eva and AI Box (1)
  • AI Deployment in Intelligent Cockpits (4): Eva and AI Box (2)
  • AI Deployment in Intelligent Cockpits (5): Super Eva+G-ASD 4.0 (1)
  • AI Deployment in Intelligent Cockpits (5): Super Eva+G-ASD 4.0 (2)
  • AI Deployment in Intelligent Chassis (1): AI Digital Chassis
  • AI Deployment in Intelligent Chassis (2): AI Power Control and Battery Management
  • 5.6 BYD
  • AI Strategy (1): Vehicle Intelligence Strategy
  • AI Strategy (2): Three Strategic Goals
  • AI Computing Power Strategy: Edge-Cloud Computing Power
  • AI Chip Strategy (1): Self-developed 4nm Xuanji A3 Chip
  • AI Chip Strategy (2): Chip Cooperation
  • AI Chip Strategy (3): BYD9000 Jointly Developed with MediaTek
  • AI Deployment Strategies for Operating Systems: BYD OS Architecture
  • AI Deployment in Intelligent Driving (1): AI Foundation Model Iteration
  • AI Deployment in Intelligent Driving (2): DiPilot (1)
  • AI Deployment in Intelligent Driving (2): DiPilot (2)
  • AI Deployment in Intelligent Driving (3): Xuanji AI Large Model (1)
  • AI Deployment in Intelligent Driving (3): Xuanji AI Large Model (2)
  • AI Deployment in Intelligent Driving (4): Integrating Deepseek R1 Large Model
  • AI Deployment in Intelligent Cockpits (1): AI Application
  • AI Deployment in Intelligent Cockpits (2): DiLink Cockpit Platform
  • AI Deployment in Intelligent Cockpits (3): AI Multimodal Interaction
  • AI Deployment in Intelligent Cockpits (4): Integrating Tongyi Large Model
  • AI Deployment in Intelligent Cockpits (5): Integrating Doubao Large Model
  • AI Deployment in Intelligent Cockpits (6): Synchronous Upgrades of Software and Hardware, Planned Application of AI-Powered Intelligent Cockpits
  • 5.7 AITO
  • AI Strategy
  • AI Data Strategy (1): Data Collection
  • AI Data Strategy (2): Data Training
  • AI Computing Power Strategy (1): Edge-Cloud Collaboration
  • AI Computing Power Strategy (2): Edge Computing Power
  • AI Chip Strategy (1): Ascend AI chip
  • AI Deployment Strategies for Operating Systems (1): HarmonyOS (1)
  • AI Deployment Strategies for Operating Systems (1): HarmonyOS (2)
  • AI Deployment Strategies for Operating Systems (2): Harmony Intelligent Mobility Alliance (HIMA)+Deepseek
  • AI Deployment in Intelligent Driving (1): AI Foundation Model Iteration
  • AI Deployment in Intelligent Driving (2): Layering of Pangu Model
  • AI Deployment in Intelligent Driving (3): MoLA (1)
  • AI Deployment in Intelligent Driving (3): MoLA (2)
  • AI Deployment in Intelligent Driving (3): MoLA (3)
  • AI Deployment in Intelligent Driving (4): Cloud World Engine of HIMA ADS 4
  • AI Deployment in Intelligent Driving (5): Automotive World Action Model of HIMA ADS 4
  • AI Deployment in Intelligent Cockpits (1): AI Cockpit Foundation Model
  • AI Deployment in Intelligent Cockpits (2): Qianwu Model (1)
  • AI Deployment in Intelligent Cockpits (2): Qianwu Model (2)
  • AI Deployment in Intelligent Cockpits (3): Xiaoyi (1)
  • AI Deployment in Intelligent Cockpits (3): Xiaoyi (2)
  • 5.8 Changan Automobile
  • AI Strategy: Beidou Dubhe Strategy (1)
  • AI Strategy: Beidou Dubhe Strategy (2)
  • AI Data Strategy
  • AI Computing Power Strategy (1): Cloud Computing Power
  • AI Computing Power Strategy (2): Edge Computing Power
  • AI Chip Strategy: Chip Cooperation
  • AI Deployment Strategies for Operating Systems (1): Tops OS
  • AI Deployment Strategies for Operating Systems (2): Integrate AI into SOA Layer
  • AI Deployment in Intelligent Driving: BEV+LLM+GoT E2E System
  • AI Deployment in Intelligent Cockpits (1): StellarWave Model
  • AI Deployment in Intelligent Cockpits (2): SDA Cockpit
  • AI Deployment in Intelligent Cockpits (3): Topspace
  • AI Deployment in Intelligent Cockpits (4): Deepal OS Intelligent Cockpit
  • AI Deployment in Intelligent Cockpits (5): Deepal AI Intelligent Cockpit
  • 5.9 BAIC
  • AI Strategy: "Baimo Huichuang" Strategy
  • AI Computing Power Strategy: Edge Computing Power
  • AI Chip Strategy: Chip Cooperation
  • AI Deployment Strategies for Operating Systems (1): AIOS
  • AI Dep