NVIDIA's outlook for 2025: Structural changes brought about by the 'AI Factory Strategy'.
NVIDIA $NVDA is often referred to as a "semiconductors company," but that understanding is incorrect. The company is evolving into a "platform company" that provides foundational technology for AI infrastructure. The AI boom is superficially discussed as a "semiconductor supply shortage," but its essence is different. It is a phenomenon arising from a structural demand excess during the fundamental reconstruction process of AI infrastructure. Particularly noteworthy is the changing demand pattern of datacenter operators. The nature of investment is transitioning from traditional "planned equipment investments" to "strategic investments for competitive advantage."
The industrial revolution of the 18th century began with a single technological innovation - the steam engine. Subsequently, with electricity and then computers, technological innovations have fundamentally changed human modes of production. Now, we are witnessing the fourth wave, the industrial revolution through digital intelligence. At the core of this new industrial revolution is the concept of an "AI Factory".
Jensen Huang, the CEO of NVIDIA $NVDA, positions this not as a mere evolution of datacenters but as a new industrial infrastructure that "manufactures" digital intelligence. What makes the AI Factory decisively different from traditional datacenters is its mode of production. Traditional software development was based on manual coding line by line by human programmers. However, in the AI Factory, machine learning models (AI) directly learn from data to generate new intelligence. This is the most fundamental paradigm shift in the history of software development. Supporting this shift is the innovation in computing architecture.
The conventional CPU-centric computing model was not optimized for parallel computation of neural networks. NVIDIA's GPU architecture has solved this challenge, making large-scale machine learning possible. Particularly, the new generation Blackwell architecture dramatically improves power efficiency, enabling the training and inference of even larger AI models. More importantly, it is the fusion of Agentic AI and Physical AI. Agentic AI refers to active AI evolved from traditional passive AI tools. It autonomously performs tasks like human colleagues rather than just computational aids. On the other hand, Physical AI understands causal relationships in the physical world and brings innovation to industrial robotics.
The economic impact of this technological innovation is clearly reflected in the third-quarter financial results. The remarkable growth of revenue of $35.1 billion (a 94% increase from the same period last year) strongly supports the effectiveness of the AI Factory strategy. Furthermore, the expected revenue of $37.5 billion for the fourth quarter indicates that this growth is accelerating. As BofA analysts point out, the resilience of demand across cloud, enterprise, and sovereign customers suggests that this transformation is structural rather than a temporary boom. In particular, the demand for Blackwell architecture is expected to significantly exceed the initial forecast of billions of dollars, according to Citi analysts.
Towards 2025, the full-scale deployment of the AI Factory model is set to begin. Analysts at Melius predict a revenue growth of 51% compared to the previous year in 2026, indicating the scale of growth. Goldman Sachs also highlights the expansion of demand for AI infrastructure, supply improvements, and margin normalization in the second half of 2025, anticipating further growth.
Furthermore, the growth rate of 'over 60% CAGR until 2027' predicted by Mizuho may actually be a conservative estimate when considering the transformation of a trillion-dollar datacenter infrastructure into AI factories.
📍Stock price forecast: Outlook from 2025.
There are various opinions on the stock price estimates for NVIDIA $NVDA at present. While Deutsche Bank offers a conservative target price of $140, Rosenblatt Securities sets an ambitious target of $220. This difference reflects the differing interpretations of how to evaluate the company's business model transformation. The revenue of $35.1 billion in the third quarter earnings (+94% year-on-year) and the expected revenue of $37.5 billion in the fourth quarter strongly suggest the feasibility of the AI Factory strategy.
2025 will be the period of full-scale deployment of Blackwell Architecture. Although temporary margin declines are inevitable, it is a strategic decision that should be actively evaluated from the perspective of securing market share and establishing a long-term revenue foundation.
Towards 2025 and beyond, more structural changes will become apparent. As the $1 trillion datacenter infrastructure transitions to AI Factories, NVIDIA's revenue structure is expected to evolve from traditional semiconductor companies to more stable and high-profit platform companies. This change will justify a higher valuation multiple.
Considering such structural changes, price levels of $160-190 by 2025 and $200 towards 2026 come into view. This is not simply an extension of revenue forecasts but based on the evaluation of the value as a platformer in the digital intelligence industrial revolution. The conventional PER of 54 times could potentially expand to around 72 times, based on past patterns. This valuation is sufficiently justified even when compared to platform companies like Amazon $AMZN and Microsoft $MSFT.
What is important is that this stock price forecast is not just based on technical analysis or extrapolation of past trends, but on the realization of the AI Factory strategy and its value creation. NVIDIA is leading a fundamental transformation in the computing industry, and the value created in the process is likely to significantly exceed the current market valuation.
The industrial revolution of the 18th century began with a single technological innovation - the steam engine. Subsequently, with electricity and then computers, technological innovations have fundamentally changed human modes of production. Now, we are witnessing the fourth wave, the industrial revolution through digital intelligence. At the core of this new industrial revolution is the concept of an "AI Factory".
Jensen Huang, the CEO of NVIDIA $NVDA, positions this not as a mere evolution of datacenters but as a new industrial infrastructure that "manufactures" digital intelligence. What makes the AI Factory decisively different from traditional datacenters is its mode of production. Traditional software development was based on manual coding line by line by human programmers. However, in the AI Factory, machine learning models (AI) directly learn from data to generate new intelligence. This is the most fundamental paradigm shift in the history of software development. Supporting this shift is the innovation in computing architecture.
The conventional CPU-centric computing model was not optimized for parallel computation of neural networks. NVIDIA's GPU architecture has solved this challenge, making large-scale machine learning possible. Particularly, the new generation Blackwell architecture dramatically improves power efficiency, enabling the training and inference of even larger AI models. More importantly, it is the fusion of Agentic AI and Physical AI. Agentic AI refers to active AI evolved from traditional passive AI tools. It autonomously performs tasks like human colleagues rather than just computational aids. On the other hand, Physical AI understands causal relationships in the physical world and brings innovation to industrial robotics.
The economic impact of this technological innovation is clearly reflected in the third-quarter financial results. The remarkable growth of revenue of $35.1 billion (a 94% increase from the same period last year) strongly supports the effectiveness of the AI Factory strategy. Furthermore, the expected revenue of $37.5 billion for the fourth quarter indicates that this growth is accelerating. As BofA analysts point out, the resilience of demand across cloud, enterprise, and sovereign customers suggests that this transformation is structural rather than a temporary boom. In particular, the demand for Blackwell architecture is expected to significantly exceed the initial forecast of billions of dollars, according to Citi analysts.
Towards 2025, the full-scale deployment of the AI Factory model is set to begin. Analysts at Melius predict a revenue growth of 51% compared to the previous year in 2026, indicating the scale of growth. Goldman Sachs also highlights the expansion of demand for AI infrastructure, supply improvements, and margin normalization in the second half of 2025, anticipating further growth.
Furthermore, the growth rate of 'over 60% CAGR until 2027' predicted by Mizuho may actually be a conservative estimate when considering the transformation of a trillion-dollar datacenter infrastructure into AI factories.
📍Stock price forecast: Outlook from 2025.
There are various opinions on the stock price estimates for NVIDIA $NVDA at present. While Deutsche Bank offers a conservative target price of $140, Rosenblatt Securities sets an ambitious target of $220. This difference reflects the differing interpretations of how to evaluate the company's business model transformation. The revenue of $35.1 billion in the third quarter earnings (+94% year-on-year) and the expected revenue of $37.5 billion in the fourth quarter strongly suggest the feasibility of the AI Factory strategy.
2025 will be the period of full-scale deployment of Blackwell Architecture. Although temporary margin declines are inevitable, it is a strategic decision that should be actively evaluated from the perspective of securing market share and establishing a long-term revenue foundation.
Towards 2025 and beyond, more structural changes will become apparent. As the $1 trillion datacenter infrastructure transitions to AI Factories, NVIDIA's revenue structure is expected to evolve from traditional semiconductor companies to more stable and high-profit platform companies. This change will justify a higher valuation multiple.
Considering such structural changes, price levels of $160-190 by 2025 and $200 towards 2026 come into view. This is not simply an extension of revenue forecasts but based on the evaluation of the value as a platformer in the digital intelligence industrial revolution. The conventional PER of 54 times could potentially expand to around 72 times, based on past patterns. This valuation is sufficiently justified even when compared to platform companies like Amazon $AMZN and Microsoft $MSFT.
What is important is that this stock price forecast is not just based on technical analysis or extrapolation of past trends, but on the realization of the AI Factory strategy and its value creation. NVIDIA is leading a fundamental transformation in the computing industry, and the value created in the process is likely to significantly exceed the current market valuation.
Disclaimer: Community is offered by Moomoo Technologies Inc. and is for educational purposes only.
Read more
Comment
Sign in to post a comment
Magnificent7 : Amazing.