在生成式 AI 應用快速擴散的當下，全球個人電腦（PC）產業正經歷自網際網路問世以來最重大的技術典範轉移。AI 運算需求從雲端資料中心向邊緣設備擴散，使得內建「神經網路處理單元」（NPU）的 AI PC 成為運算的核心硬體基礎，並對上游半導體廠的競爭格局產生根本性的衝擊。然而，現有策略管理文獻對半導體競爭的分析多採取供給端視角，較少引入下游 OEM 品牌廠的採購決策作為策略成效的實證驗證：「技術推力」能否真正轉化為下游採購的「需求拉力」，仍是既有研究尚未充分回應的問題。

本研究以司徒達賢（2024）的策略形態分析法為核心分析工具，並整合資源基礎觀點（Barney, 1991）與動態能力理論（Teece, Pisano, & Shuen, 1997），對 Intel、AMD 與 NVIDIA 三家半導體巨頭在 AI PC 發展下的競爭策略進行多重個案研究（Yin, 2018）。本研究最重要的方法論，在於引入全球領先 PC 品牌廠（OEM A）2024 至 2026 年的機密內部產品發展藍圖，作為驗證上游策略形態能否轉化為下游採購訂單的實證依據。

研究發現可歸納為三項相互支撐的核心觀察。第一，三家晶片廠在 AI PC 時代展現出顯著差異化的策略形態——Intel 採取以歷史 x86 生態系為基礎的「廣泛差異化」、AMD 採取結合小晶片彈性與腳位沿用的「最佳成本混合策略」、NVIDIA 則採取以 CUDA 生態系與 N1/N1X SoC 為核心的「集中差異化」。第二，OEM A 的機密藍圖實證顯示，三家廠商的策略形態確實在採購決策中獲得呼應，但呼應方式並非「最強策略形態全面勝出」，而是「各廠策略形態與特定市場區隔需求對齊」的多供應商分區配置結構——Intel 主導商用主流與極致輕薄、AMD 主導消費旗艦與部分主流商用、NVIDIA 切入頂級旗艦的「測試點」配置。第三，在 AI PC 典範轉移的當前階段，Fabless 模式（AMD/NVIDIA）具備較 IDM 模式（Intel）更強的短期資源重配置敏捷性；IDM 模式的長期競爭力恢復，將取決於 Intel 18A 製程能否達到台積電同等水準，以及 Intel Foundry 能否建立足夠的外部客戶基礎。

本研究在理論貢獻上，將司徒達賢的本土策略管理理論框架延伸應用至全球半導體供應鏈情境，並透過引入 OEM 機密藍圖，填補了既有研究在「供給端策略分析」與「需求端買方驗證」之間的缺口；同時也對 IDM vs. Fabless 比較文獻提供了 AI 時代的實證修正。在實務意涵上，本研究依循外部觀察者立場，歸納三家晶片廠既有策略形態邏輯所指向的調適方向：Intel 的策略張力指向 Intel 18A 製程良率的提升與 Intel Foundry 外部客戶基礎的建立；AMD 指向 Chiplet IP 優勢的深化與 ROCm 軟體生態的強化；NVIDIA 則指向 N1/N1X SoC 與 x86 RTX SoC 雙軌定位的釐清，以降低 OEM 採購不確定性。另一方面，本研究對 OEM 品牌廠在多供應商配置與供應鏈韌性管理上提出三項具體建議，期能為 AI 典範轉移時代的半導體競爭策略提供學術與實務參考。

Amid the rapid diffusion of generative AI, the global personal computer (PC) industry is undergoing its most significant technological paradigm shift since the rise of the internet. As AI computing demand migrates from cloud data centers to edge devices, AI PCs equipped with dedicated Neural Processing Units (NPUs) have emerged as the core hardware foundation of next-generation computing, fundamentally reshaping the competitive landscape of upstream semiconductor firms. However, existing strategic management literature on semiconductor competition has predominantly adopted a supply-side perspective, with limited integration of downstream OEM procurement decisions as an empirical validation mechanism for strategic effectiveness. Whether "technology push" from upstream chipmakers truly translates into "demand pull" from downstream procurement remains an underexplored question in the literature.

This study employs Seetoo's (2024) Strategic Configuration Analysis as its core analytical framework, integrated with the Resource-Based View (Barney, 1991) and Dynamic Capabilities theory (Teece, Pisano, & Shuen, 1997), to conduct a multiple case study (Yin, 2018) of Intel, AMD, and NVIDIA in the context of the AI PC paradigm shift. The most significant methodological contribution of this study lies in incorporating the confidential internal product roadmaps (2024–2026) of one of the world's top PC brand vendors (OEM A) as objective empirical evidence to verify whether upstream strategic configurations translate into downstream procurement orders.

The findings yield three mutually reinforcing core observations. First, the three semiconductor firms exhibit distinctly differentiated strategic configurations in the AI PC era: Intel pursues "Broad Differentiation" anchored in its legacy x86 ecosystem; AMD adopts a "Best-Cost Provider" strategy combining chiplet flexibility with socket-compatible designs; NVIDIA takes a "Focused Differentiation" approach centered on its CUDA ecosystem and N1/N1X SoC. Second, OEM A's confidential roadmap empirically demonstrates that these strategic configurations are indeed reflected in procurement decisions—but not as a "winner-takes-all" outcome. Rather, the procurement pattern reveals a "segment-aligned multi-sourcing structure": Intel dominates corporate mainstream and ultra-thin segments, AMD dominates consumer flagship and parts of mainstream commercial segments, while NVIDIA enters the premium flagship segment as a "test point" deployment. Third, in the current phase of the AI PC paradigm shift, the Fabless model (AMD/NVIDIA) demonstrates greater short-term resource reconfiguration agility than the IDM model (Intel); the long-term competitive recovery of the IDM model will depend on whether Intel's 18A process can match TSMC's standards and whether Intel Foundry can establish a sufficient external customer base.

This study contributes theoretically by extending Seetoo's locally-developed strategic management framework to the global semiconductor supply chain context, and by filling the knowledge gap between supply-side strategic analysis and demand-side buyer validation through the introduction of confidential OEM roadmap data. It further provides an empirical revision to the IDM vs. Fabless comparative literature in the AI era. From a practical standpoint, rather than prescribing strategies to the firms, the study adopts an external-observer stance and identifies the adjustment directions implied by each firm's existing strategic configuration: for Intel, the strategic tension points toward improving Intel 18A process yield while establishing an external Intel Foundry customer base; for AMD, toward deepening its chiplet IP advantages and strengthening the ROCm software ecosystem; and for NVIDIA, toward clarifying the dual-path positioning between its N1/N1X SoC and x86 RTX SoC platforms to reduce OEM procurement uncertainty. The study additionally offers three concrete recommendations for OEM brand vendors regarding multi-sourcing configuration and supply chain resilience management. These insights aim to provide both academic and practical references for understanding semiconductor competitive strategy in the era of the AI paradigm shift.

ㄧ、中文文獻
司徒達賢 (2016)。策略管理新論：觀念架構與分析方法(三版)。台北：智勝文化事業有限公司。
司徒達賢 (2024)。策略管理講義。台北：國立政治大學企業管理研究所。https://ba.nccu.edu.tw/zh_tw/download/SEETOO2025ST

二、英文文獻
Adobe. (2026). Agentic AI and the future of human-computer interaction. Author.
AInvest. (2025, December 31). Is Intel's foundry turnaround finally gaining traction? AInvest. https://www.ainvest.com/news/intel-foundry-turnaround-finally-gaining-traction-2512
AInvest. (2026, January). Intel's 18A: Assessing the S-curve of a foundry infrastructure play. AInvest.https://www.ainvest.com/news/intel-18a-assessing-curve-foundry-infrastructure-play-2601/
Albanesius, C. (2025). From Zen 7 to Medusa: AMD tips next-gen chips in new roadmaps. PCMag. https://www.pcmag.com/news/from-zen-7-to-medusa-amd-tips-next-gen-chips-in-new-roadmaps
AMD. 2024. Client processor platform roadmap 2024–2026 [Confidential internal roadmap provided to OEM A].
AMD. 2025. Evaluating the future of the AI PC: Accelerate your organization’s AI strategy. Author.
AMD. 2026. AMD expands AI leadership across client, graphics, and software with new Ryzen, Ryzen AI, and AMD ROCm announcements at CES 2026 [Press release]. Santa Clara, CA: Author.
Aminext. (2025, September 22). Why Nvidia's true moat isn't chips, but CUDA: An investor's guide to the ecosystem wars. Aminext. https://www.aminext.blog/en/post/why-nvidia-s-true-moat-isn-t-chips-but-cuda-an-investor-s-guide-to-the-ecosystem-wars
Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120.
Besard, T., Foket, C., & De Sutter, B. (2018). Effective extensible programming: Unleashing Julia on GPUs. IEEE Transactions on Parallel and Distributed Systems, 30(4), 827–841.
Bonen, N., Gihon, A., Polishuk, L., Aizik, Y., Okunev, Y., Kurts, T., & Bashyam, N. (2025). Lunar Lake an Intel mobile processor: SoC architecture overview (2024). IEEE Micro, 45(3), 15–21. https://doi.org/10.1109/MM.2025.3558407
Bower, J. L. (1970). Managing the resource allocation process: A study of corporate planning and investment. Harvard Business School Press.
Chang, S.-J., & Matsumoto, Y. (2022). Dynamic resource redeployment in global semiconductor firms. Strategic Management Journal, 43(2), 237–265.
Climent, R. C., Haftor, D. M., & Staniewski, M. W. (2024). AI-enabled business models for competitive advantage. Journal of Innovation & Knowledge, 9, 100532.
Costantino, N., & Pellegrino, R. (2010). Choosing between single and multiple sourcing based on supplier default risk: A real options approach. Journal of Purchasing and Supply Management, 16(1), 27–40.
Dally, W. J., Keckler, S. W., & Kirk, D. B. (2021). Evolution of the graphics processing unit (GPU). IEEE Micro, 41(6), 42–51.
Duprey, R. 2026. Intel lands Musk’s $25 billion Terafab: A billion-dollar foundry win in the making? 24/7 Wall St. https://247wallst.com/investing/2026/04/07/intel-lands-musks-25-billion-terafab-a-billion-dollar-foundry-win-in-the-making/
Fainshmidt, S., Wenger, L., Pezeshkan, A., & Mallon, M. R. (2019). When do dynamic capabilities lead to competitive advantage? The importance of strategic fit. Journal of Management Studies, 56(4), 758–787.
Finterra. (2026, February 5). AMD: The architecture of an AI powerhouse (2026 deep dive). Financial Content. https://markets.financialcontent.com/wral/article/finterra-2026-2-5-amd-the-architecture-of-an-ai-powerhouse-2026-deep-dive
Gartner. (2025). Gartner says AI PCs will represent 31% of worldwide PC market by the end of 2025. Author.
Gartner. (2026). Gartner says surging memory costs will reduce global PC and smartphone shipments in 2026. Author.
Gawer, A., & Cusumano, M. A. (2002). Platform leadership: How Intel, Microsoft, and Cisco drive industry innovation. Harvard Business School Press.
Goettler, R. L., & Gordon, B. R. (2011). Does AMD spur Intel to innovate more? Journal of Political Economy, 119(6), 1141–1200.
Gomes, W., Morgan, S., Phelps, B., Wilson, T., & Hallnor, E. (2022). Meteor Lake and Arrow Lake: Intel next-gen 3D client architecture platform with Foveros. Paper presented at the 2022 IEEE Hot Chips 34 Symposium (HCS). https://doi.org/10.1109/HCS55958.2022.9895532
Gupta, A., & Sharma, B. (2021). Changing trends in computer architecture: A comprehensive analysis of ARM and x86 processors. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 7(3), 619–631.
Hennessy, J. L., & Patterson, D. A. (2017). Computer architecture: A quantitative approach (6th ed.). Morgan Kaufmann.
Hong, Y., & Kim, D. (2025). Performance and efficiency gains of NPU-based servers over GPUs for AI model inference. Systems, 13(9), 797.
Hyperframe Research. (2025, December 24). Can Google's TorchTPU eventually bridge NVIDIA's CUDA moat? Hyperframe Research https://hyperframeresearch.com/2025/12/24/can-googles-torchtpu-eventually-bridge-nvidias-cuda-moat/
IDC. (2026). Memory shortage crisis hits PCs and smartphones. Author.
Intel Corporation. (2024). Client platform roadmap 2024–2026 [Confidential internal roadmap provided to OEM A].
Intel Corporation. (2026). Intel's AI strategy, offerings, and differentiation in the AI PC era. Author.
Jiang, W. (2025). Analysis of differences in financial management across enterprises: A case study of the semiconductor industry. Proceedings of ICFTBA 2025 Symposium.
Khan, F. H., Pasha, M. A., & Masud, S. (2021). Advancements in microprocessor architecture for ubiquitous AI—An overview on history, evolution, and upcoming challenges in AI implementation. Micromachines, 12(6), 665.
Khan, S. M., Peterson, D., & Mann, A. (2021). The semiconductor supply chain: Assessing national competitiveness. Center for Security and Emerging Technology.
Klover.ai. (2025, July 25). NVIDIA AI strategy: Analysis of sustained dominance in AI. Klover.ai. https://www.klover.ai/nvidia-ai-strategy-analysis-sustained-dominance-ai/
Lee, J., Suh, T., Roy, D., & Baucus, M. (2019). Emerging technology and business model innovation: The case of artificial intelligence. Journal of Open Innovation: Technology, Market, and Complexity, 5(3), 44.
Leong, J. (2022). Nvidia and the great East-West semiconductor game (Case No. UVA-S-0374). University of Virginia Darden School Foundation.
Lincoln, Y. S., & Guba, E. G. (1985). But is it rigorous? Trustworthiness and authenticity in naturalistic evaluation. In D. D. Williams (Ed.), Naturalistic evaluation (pp. 73–84). Jossey-Bass.
Macher, J. T., & Mowery, D. C. (2009). Measuring dynamic capabilities: Practices and performance in semiconductor manufacturing. British Journal of Management, 20, S41–S62.
Malkin, A., & He, T. (2024). The geoeconomics of global semiconductor value chains: Extraterritoriality and the US-China technology rivalry. Review of International Political Economy, 31(2), 674–699.
MarketsandMarkets. (2025). AI PC market by product, operating system, compute architecture, compute type, price, end user, and region – Global forecast to 2031. Author.
Martin, D. (2026, February 11). Intel's supply issues helped AMD grab record-high CPU market share: Researcher. CRN. https://www.crn.com/news/components-peripherals/2026/intel-s-supply-issues-helped-amd-grab-record-high-cpu-market-share-researcher
Mercury Research. 2026. 2025 Q4 x86 processor market share report [As reported by iThome]. https://www.ithome.com.tw/pr/173938
Miller, D. (1986). Configurations of strategy and structure: Towards a synthesis. Strategic Management Journal, 7(3), 233–249.
Mintzberg, H. (1979). The structuring of organizations. Prentice-Hall.
Morgan Stanley Research. (2026, February 11). Mapping AI's rate of change: Where the rubber meets the road. New York, NY: Author.
NotebookCheck. (2020, February 28). AMD says chiplet design can cut costs by more than half. NotebookCheck.https://www.notebookcheck.net/AMD-says-chiplet-design-can-cut-costs-by-more-than-half.454676.0.html
NVIDIA Corporation. (2024). GeForce and SoC platform roadmap 2024–2026 [Confidential internal roadmap provided to OEM A].
NVIDIA Corporation. (2025). NVIDIA GTC 2025 keynote [Conference presentation]. Author.
NVIDIA Corporation. (2026). NVIDIA GTC 2026 keynote [Conference presentation]. https://www.nvidia.com/gtc/keynote/
OEM A. (2024). AI PC product planning roadmap 2024–2027 [Unpublished internal planning document].
OEM A. (2025). AI PC product planning roadmap 2025–2028 [Unpublished internal planning document].
Oplexa. (2026). NVIDIA GTC 2026: 5 big announcements investors should watch. Oplexa. https://oplexa.com/nvidia-gtc-2026-announcements-investors/
PC Gamer. (2025). AMD's CPU division is booming as CEO Dr. Lisa Su says sales 'far exceeded my expectations'. PC Gamer. https://www.pcgamer.com/hardware/processors/amds-cpu-division-is-booming-as-ceo-dr-lisa-su-says-sales-far-exceeded-my-expectations/
Peccerillo, B., Mannino, M., Mondelli, A., & Bartolini, S. (2022). A survey on hardware accelerators: Taxonomy, trends, challenges, and perspectives. Journal of Systems Architecture, 129, 102561.
Peteraf, M. A. (1993). The cornerstones of competitive advantage: A resource-based view. Strategic Management Journal, 14(3), 179–191.
Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries and competitors. Free Press.
Qualcomm. (2024). Heterogeneous computing: The architecture of edge AI and NPU performance. Author.
Reed, R., & DeFillippi, R. J. (1990). Causal ambiguity, barriers to imitation, and sustainable competitive advantage. Academy of Management Review, 15(1), 88–102.
Regi, G., Kunnappally, J. J., & Yunus, R. (2024). Comparative analysis of energy efficiency between ARM and x86 architectures in mobile devices and its implications for human-computer interaction. Kristu Jayanti Journal of Computational Sciences, 4, 54–67.
Rumelt, R. P. (1984). Toward a strategic theory of the firm. In R. B. Lamb (Ed.), Competitive strategic management (pp. 556–570). Prentice-Hall.
Shen, R. (2023). Intel's acquisition of Tower Semiconductor: Will foundry be the future of Intel. Highlights in Business, Economics and Management, 7, 335–341.
Silicon Analysts. (2026). Industry analysis – Semiconductor market intelligence. https://siliconanalysts.com/analysis
Straits Research. (2025). AI PC market size, share, trends and growth analysis. Author.
Suggs, D., Subramony, M., & Bouvier, D. (2019). The AMD "Zen 2" processor. Hot Chips 31 Symposium, 1–20.
Tan, T. (2024). Business analysis: A comparative analysis of AMD and NVIDIA. Proceedings of ICFTBA 2024 Workshop: Human Capital Management in a Post-Covid World: Emerging Trends and Workplace Strategies.
Tang, Y.-E., & Liou, F.-M. (2010). Does firm performance reveal its own causes? The role of Bayesian inference. Strategic Management Journal, 31(1), 39–57.
TechInsights. (2024). Analysis: AI PCs are a bright spot in the sagging Intel portfolio. Author.
TechInsights. (2026, March). Narrative: AI PC processor market share 2025. TechInsights. https://www.techinsights.com/blog/narrative-ai-pc-processor-market-share-2025
TechPowerUp. (2026). NVIDIA's N1X CPU hits new roadblock: Launch pushed to late 2026. TechPowerUp. https://www.techpowerup.com/338938/nvidias-n1x-cpu-hits-new-roadblock-launch-pushed-to-late-2026
Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509–533.
Teece, D. J. (2018). Business models and dynamic capabilities. Long Range Planning, 51(1), 40–49.
The Next Platform. (2021, June 9). AMD on why chiplets—and why now. The Next Platform. https://www.nextplatform.com/2021/06/09/amd-on-why-chiplets-and-why-now/
Tom's Hardware. (2026). Alleged images of the long-awaited Nvidia N1/N1X SoC surface on laptop motherboard. Tom's Hardware. https://www.tomshardware.com/pc-components/cpus/alleged-images-nvidia-n1-n1x-soc-surface
Trefis. (2026). Advanced Micro Devices (AMD) market analysis and server revenue share. Trefis. https://www.trefis.com/data/companies/AMD?source=forbes&from=AMD-2026-01-27
Treleven, M., & Schweikhart, S. B. (1988). A risk/benefit analysis of sourcing strategies: Single vs. multiple sourcing. Journal of Operations Management, 7(4), 93–114.
Vargo, S. L., & Lusch, R. F. (2008). Service-dominant logic: Continuing the evolution. Journal of the Academy of Marketing Science, 36(1), 1–10.
Venkatraman, N. (1989). The concept of fit in strategy research: Toward verbal and statistical correspondence. Academy of Management Review, 14(3), 423–444.
Wadhwani, P., & Jaiswal, A. (2025). AI computer market size, share & industry analysis, 2025–2034. Global Market Insights Inc.
Wang, C.-H., & Lin, H.-C. (2021). Competitive substitution and technological diffusion for semiconductor foundry firms. Advanced Engineering Informatics, 48, 101254.
Wccftech. (2026, March 6). NVIDIA secured 94% GPU market share in Q4 25, AIB market gripped by rising prices. Wccftech. https://wccftech.com/nvidia-secured-94-percent-aib-gpu-market-share-q4-25/
WorkBoat. (2026, January 14). The AMD transformation: From x86 underdog to AI systems architect. Financial Content. https://markets.financialcontent.com/workboat/article/finterra-2026-1-14-the-amd-transformation-from-x86-underdog-to-ai-systems-architect
Yin, R. K. (2018). Case study research and applications: Design and methods (6th ed.). SAGE Publications.