Last week officially announced entry into x86, now a hundred billion dollar gamble on OpenAI! NVIDIA's "AI computing power myth" reaches its peak

There is no doubt that since the beginning of 2023, when ChatGPT became popular worldwide, prompting enterprises and government agencies to invest heavily in AI, "AI chip giant" NVIDIA (NVDA.US) has been cashing in on this unprecedented AI wave. The incredibly strong revenue and operating profit achieved in recent years have provided a solid cash flow foundation. Therefore, at a time when global AI infrastructure is booming and AI applications like DeepSeek, ChatGPT, and Claude are becoming increasingly powerful and accelerating their penetration into various industries, NVIDIA has invested heavily to initiate the "AI ecosystem integration phase." The moat of this AI chip giant is beginning to transform into a "fortress," and the "NVIDIA AI Empire" seems unshakeable. According to top analysts on Wall Street, the AI integration phase that NVIDIA is entering mainly aims to strengthen its market dominance in AI computing infrastructure and expand its moat through strategic investments, collaborations, and potential acquisitions. For example, Zhitong Finance APP reported that NVIDIA announced on Monday local time that it would invest up to $100 billion in leading AI company OpenAI to support the construction of next-generation data centers and AI infrastructure. The two parties plan to jointly build at least 10 gigawatts (GW) of computing power scale super AI data centers, equivalent to the total output power of 10 traditional nuclear power plants. This unprecedented collaboration highlights the unparalleled demand for AI computing infrastructure brought about by the accelerated penetration of generative AI tools like ChatGPT into global industries. In addition, another significant news regarding NVIDIA's AI integration is that NVIDIA announced last week a strategic partnership with its long-time competitor Intel. This strategic cooperation has significantly expanded NVIDIA's total addressable market (TAM), further deepening NVIDIA's control over the AI ecosystem and integrating NVIDIA's GPU acceleration technology into the vast personal consumer and enterprise server CPU market covered by Intel. As of Monday's market close, NVIDIA's stock price reached a new all-time high, hovering around $183, with a market capitalization of approximately $4.46 trillion, reflecting a 36% increase this year. Since the year’s lowest point in early April, when Trump's aggressive tariff policies triggered a global stock market crash, the increase has been an astonishing 95%. Top institutions on Wall Street believe that NVIDIA's stock price's trend of repeatedly hitting new highs is far from over. Wall Street analysts have continuously raised their 12-month target price for NVIDIA, and the latest average target price from Wall Street implies that NVIDIA's total market capitalization will break the $5 trillion milestone within a year. Although NVIDIA's valuation multiples are at a premium level, its sustained strong growth trajectory and profit momentum will support the "strong buy" rating given by Wall Street analysts, indicating further upward potential. As the "highest market capitalization company in the world," NVIDIA is regarded as the "leader" of the global AI computing industry chain. Therefore, for institutional and retail investors, NVIDIA's strong upward momentum, which continues to set new highs, signifies that this round of "super bull market" in the global AI computing industry chain is far from over, and this industry chain will remain the most favored investment sector for global capital in the near future. ** It is under the epic stock price surge of large tech giants such as NVIDIA, Meta, Google, Oracle, TSMC, and Broadcom, as well as the leaders in the AI computing power industry chain, that an unprecedented AI investment boom has swept through the U.S. stock market and global stock markets. This has driven the S&P 500 index and the global benchmark index—MSCI World Index to surge significantly since April, recently setting new historical highs.** Recently, global prices for DRAM and NAND series storage products have skyrocketed, and the global cloud computing giant Oracle, with its contract reserves exceeding market expectations at $455 billion, along with Broadcom, the global AI ASIC chip "superpower," have recently announced strong performance and future outlooks that significantly reinforce the "long-term bull market narrative" for AI GPU, ASIC, and HBM and other AI computing power infrastructure sectors. The AI computing power demand driven by generative AI applications and AI agents at the inference end can be described as "starry seas," expected to drive the artificial intelligence computing power infrastructure market to continue showing exponential growth. The "AI inference system" is also considered by Jensen Huang to be the largest source of future revenue for NVIDIA. "What we see in AI infrastructure construction and the entire tech sector can almost be said to be driven by semiconductor-related stocks," said Wayne Kaufman, chief market analyst at Phoenix Financial Services. He added that Oracle's recent strong financial report and the over $10 billion AI infrastructure deal between Microsoft and Nebius Group NV are both positive signals for the global chip sector. "Companies are placing very long-term AI infrastructure orders, indicating that they feel they might be falling far behind in meeting the computing power infrastructure needed for artificial intelligence," Kaufman stated in an interview. "This means that the surge in semiconductors/chips looks very sustainable, even if some stocks may be overbought in the short term, chip giants like NVIDIA and TSMC will still be the biggest winners." According to Wall Street investment giants Loop Capital and Wedbush, the global AI infrastructure investment wave centered on AI computing hardware is far from over; it is only at the beginning. Driven by an unprecedented "AI computing power demand storm," this round of AI investment is expected to reach a scale of $2 trillion. NVIDIA CEO Jensen Huang even predicts that by 2030, AI infrastructure spending will reach $3 trillion to $4 trillion, with the scale and scope of projects providing NVIDIA with significant long-term growth opportunities. NVIDIA Moving Towards the AI Integration Stage, Its Moat Gradually Transitions to a "Super Fortress" For investors, one of the best times to increase stakes in a tech giant is when it clearly enters an integration phase and takes substantial actions to dominate its core market. At this stage, companies are best positioned to comprehensively consolidate their moats and minimize competitive risks, turning future innovations into compounding growth advantages for shareholders. According to top Wall Street investment institutions like Bank of America and Morgan Stanley, when such companies trade at reasonable prices and valuations, it is an obvious choice to buy on dips and hold long-term NVIDIA is a prime example of this type of company. In fact, the chip giant seems to be solidifying its moat to an extent that exceeds expectations, making it nearly impossible for even well-funded competitors to shake the company's existing "dominant position that can be described as a monopoly in AI chips." NVIDIA plans to invest approximately $100 billion in OpenAI, intending to advance with a non-controlling equity + supply structure, aiming for phased delivery of more than 10GW of NVIDIA AI computing power systems (based on the next-generation AI computing architecture Vera Rubin) starting in the second half of 2026. Combined with the previous $300 billion/5-year AI computing infrastructure contract signed between OpenAI and cloud computing giant Oracle (based on the Stargate project expansion), the cloud AI computing capacity and capital closed loop formed by NVIDIA, OpenAI, and Oracle is expected to continue to spill flagship AI applications like ChatGPT into broader markets through "capability as a service." The aforementioned "top three parties" can be said to weld together the closed loop of "capital, AI chips, and cloud computing infrastructure," significantly strengthening their moats and upstream profitability in the short term. The widespread diffusion of AI for all will subsequently occur through the cloud computing services/AI training and inference cloud computing platforms dominated by these three parties. This tripartite game is, for NVIDIA, a way to enhance visibility and bargaining power through equity returns + long-term delivery locking in demand growth curves; at the same time, it packages the AI supercomputing system and network (NVLink, liquid cooling, GB200/Blackwell) into an "integrated platform" output, further strengthening the AI infrastructure ecosystem centered around NVIDIA's AI computing clusters. According to estimates from Wall Street financial giant Bank of America, this investment is expected to bring NVIDIA up to $500 billion in high returns. "This cooperation includes a letter of intent, and NVIDIA will participate in the construction of at least 10 gigawatts of systems starting in the second half of 2026 with the Vera Rubin super AI computing platform. We estimate that over time, this could bring about $300 billion to $500 billion in project value, or about 3-5 times the investment return," wrote Bank of America analyst Vivek Arya in a report to clients. Additionally, NVIDIA's recent collaboration with Intel is clearly a signal that its integration strategy is being implemented. The result will be a significantly expanded total addressable market (TAM), deeper penetration, stronger AI ecosystem control, and a more solidified position for NVIDIA in both consumer and enterprise markets. Coupled with NVIDIA's strategic mergers and acquisitions in recent years, these initiatives enable the company to accelerate value creation for shareholders, and given NVIDIA's current valuation multiples, the stock appears highly attractive. Last Thursday, NVIDIA announced that it would collaborate with Intel to jointly develop multi-generation customized data center x86 architecture CPUs + NVIDIA AI GPU clusters and AI PC series products. Similarly, NVIDIA mentioned that it would directly integrate its GPU chiplets into Intel's x86 consumer-grade CPUs Intel will create optimized custom x86 CPUs for NVIDIA's AI infrastructure in data centers. More interestingly, NVIDIA has invested approximately $5 billion in Intel, indicating that the strategic partnership between the two has transcended mere product collaboration. From NVIDIA's perspective, it is clear that the company is shifting its focus towards the integration phase. For years, NVIDIA has established a dominant position in AI GPU-accelerated computing with its cutting-edge GPU architecture and CUDA software ecosystem. With the wave of AI prosperity, NVIDIA's GPUs have transitioned from being "niche accelerators" aimed at gaming and graphics to becoming the super core of the most important technological trend reshaping modern computing. Therefore, for some of the most critical hyperscale customers, NVIDIA's GPU architecture and CUDA have become the default engine driving large language models, generative AI platforms, and enterprise-level machine learning. Today, with its latest initiatives, it is evident that NVIDIA is seeking to strengthen its leadership position in the AI infrastructure market to fully dominate its field. In addition to strategic investments in OpenAI and Intel, the company has undertaken a series of strategic acquisitions in recent years: acquiring Run:ai, Deci, and OmniML in the AI software space; investing in Cumulus Networks and Mellanox in the networking sector; and acquiring the Israeli startup Excelero Storage in 2022, thereby making progress in high-performance storage. Overall, NVIDIA's strategy appears to exhibit a clear pattern: striving to integrate every layer of the AI computing stack. From this perspective, the recent collaborations with OpenAI and Intel align perfectly with this broader integration "script," which are undoubtedly key steps in further solidifying NVIDIA as a critical pillar of AI infrastructure. In fact, especially with the help of Intel, NVIDIA seems to have delineated a clear path into the mainstream consumer computing space. Traditionally, the company's core advantage has stemmed from its GPU architecture foundation and CUDA. However, by directly integrating GPU chips with Intel CPUs, NVIDIA will extend its reach into the mainstream PC lightweight computing market—a field NVIDIA has previously not ventured much into. As a result, NVIDIA will expand its focus beyond high-end workstations and gaming consoles, leaving an "NVIDIA+Intel mark" on laptops and everyday PCs used by millions of users. This undoubtedly signifies a significant expansion of Total Addressable Market (TAM), deeper penetration capabilities, and the potential formation of sustainable cash flows. Moreover, the new strategic collaboration between NVIDIA and Intel has far-reaching implications in the realm of large AI data centers and enterprise computing—x86 remains the dominant CPU architecture in these fields. By directly and simply integrating NVLink with Intel CPUs, the company effectively eliminates a major bottleneck in large-scale AI deployments. Enterprises that have already heavily relied on x86 infrastructure may seamlessly integrate NVIDIA's GPU computing clusters, thereby unlocking faster data center computing architecture performance and lower latency than relying solely on PCIe connections. As this integration process continues to evolve, and CUDA is firmly embedded in enterprise AI workloads, competitors must force customers to abandon existing x86 workflows or the enterprise-level ChatGPT that has already been updated and iterated on CUDA and x86 architecture in order to fully replace NVIDIA. This largely reinforces NVIDIA's "moat" into an impregnable "super fortress," providing it with a longer runway for compound growth. Currently, NVIDIA is clearly doubling down on integrating the AI ecosystem, building a "competitive advantage parameter" that makes it "difficult to reach" even for the most capable disruptors. In contrast, through ecosystem dominance, deep integration, and the resulting strong user stickiness, NVIDIA is exponentially widening the gap between itself and any "AI computing infrastructure chasers." NVIDIA Sets Sail! Launching an Assault on the $5 Trillion Market Cap Super Threshold Even with a 95% increase since the low in April, NVIDIA still appears equally attractive today among the 61 large-scale peers in the global semiconductor industry. Although the company's stock has an EV/EBITDA of about 43 times, which may seem expensive, especially when the median in the tech sector of the U.S. stock market is only about 19 times, the growth expectation path it offers, as seen in the above chart, fully justifies this premium multiple. NVIDIA's growth expectations are also the most outstanding among global tech stocks. Regarding NVIDIA's expected price-to-earnings ratio (forward P/E [GAAP]) of 40 times, it can similarly be said that it is higher than the tech sector's median of 33 times. However, when we look at the company's earnings trend over the past two and a half years, it becomes clear that this premium is indeed completely reasonable. Shareholders can clearly benefit directly from NVIDIA's compound growth trajectory by hopping on this value-creating train. There is a very clear connection between the company's stock price and the earnings it generates. In terms of NVIDIA's current integration phase, this is undoubtedly a strong entry opportunity, as the company's trailing price-to-earnings ratio (trailing P/E ratio [GAAP]) of 50 times is significantly lower than its historical average valuation over the past five years—up to 82 times trailing P/E. In fact, considering that the company is moving towards integration as mentioned above, its earnings creation potential is expected to achieve significant and sustained expansion in the future, driven by multiple synergistic factors. For example, NVIDIA's integration with Intel allows it to achieve higher profit margins from a proprietary CPU-GPU joint solution ecosystem that competitors cannot easily replicate. Similarly, expanding into both data centers and consumer-grade PC markets will significantly enlarge the company's Total Addressable Market (TAM), allowing it to dilute fixed costs over a larger revenue base and further drive its profitability trajectory.** However, for investors, it is important to note that given NVIDIA's evident efforts to control the entire computing stack (GPU, networking, storage, software frameworks, and now integrated CPU-GPU solutions), it faces higher execution risks compared to most companies that focus on narrower strategies. Considering the numerous "active components" and many factors beyond NVIDIA's direct control, each layer introduces dependencies on external partners and conditions. Issues such as software stack delays, supply chain coordination hiccups, or any errors could cascade through the entire stack and impact NVIDIA's performance. With recent developments, Wall Street analysts believe that NVIDIA's moat is clearly stronger than ever, covering GPU computing clusters, networking, storage, software frameworks, and now integrated CPU-GPU solutions. In fact, as NVIDIA continues to advance its AI ecosystem integration, the likelihood of competitors catching up to NVIDIA decreases, and the space for potential disruptors to pose a threat becomes smaller. Data compiled by TIPRANKS shows that most Wall Street analysts have recently set their 12-month target price for NVIDIA above $210, indicating their bets that NVIDIA's market value will surpass the $5 trillion mark. KeyBanc has raised its target price for NVIDIA to $215, Bank of America has increased it to $235, while TD Cowen and Cantor Fitzgerald have set target prices as high as $235 and $240, respectively. Craig-Hallum has significantly raised NVIDIA's target price from the previous $195 level to $245. Additionally, Loop Capital, a major player on Wall Street, believes that NVIDIA's market value is expected to exceed $6 trillion, assigning a target price of $250, the highest on Wall Street. As breakthrough AI applications like AI agents penetrate various industries globally, leading to an enormous demand for "AI inference computing power," the future prospects for AI computing infrastructure fields such as AI chips, HBM storage systems, enterprise-grade SSDs, and high-performance networking and power equipment will remain vast. The urgent need for companies to improve efficiency and reduce operational costs has recently propelled the widespread application of two core categories of AI application software—generative AI applications and AI agents—significantly. The emergence of AI agents signifies that artificial intelligence is evolving from an information assistance tool into a highly intelligent productivity tool. AI agents represented by OpenAI Deep Research and Manus can automate repetitive tasks, perform big data analysis and summarization based on incredibly powerful AI large models, and provide real-time monitoring insights and reports, making appropriate decisions in extremely complex situations in a very short time, thereby improving operational efficiency for enterprises. The logic of enhancing personal learning and work efficiency is fundamentally similar. AI agents can also efficiently participate in large-scale projects across various fields globally, from blueprint planning to implementation, significantly accelerating project progress. Loop Capital, which is bullish on NVIDIA's stock price reaching $250, stated in its research report: “Our research indicates that we are entering the next wave of 'golden tide' in AI applications, and NVIDIA is still at the forefront of a crucial phase where the next wave of demand is far stronger than expected.” The institution predicts that by 2028, global cloud computing giants + technology companies + sovereign AI will cumulatively spend approximately $2 trillion on computing infrastructure centered around NVIDIA AI GPUs