Wall Street dreams of a "AI super bull market" sequel! From the strongest profits in the S&P 500 in five years to the crazy rise of AI computing power beyond the Rubin epic GPU.

As the market anticipates the imminent long-term peace agreement between the United States and Iran under the mediation of Pakistan - transforming the previously fragile ceasefire agreement into a lasting peace agreement, this increasingly positive prospect has boosted Wall Street's bullish sentiment, driving one of the benchmark stock indexes in the United States, the S&P 500 index, towards its longest period of consecutive weekly gains since December 2023. According to Wall Street analysts, the Mag 7 (seven major tech giants in the US) and a multitude of AI computing power industry giants leading the US super bull market are not ending, but instead transitioning from the "AI valuation narrative" to the "AI profit realization + easing geopolitical risks + profit breadth expansion of the computing power value chain" to a more stable stage. The so-called "seven major tech giants," or the "Magnificent Seven" (also known as "Mag 7"), as viewed by Wall Street analysts, include: Apple, Microsoft, Google's parent company Alphabet, Tesla, Nvidia, Amazon, and Facebook's parent company Meta Platforms. They are the core driving force behind the record highs of the S&P 500 index and are considered by top Wall Street investment firms to be a combination capable of bringing huge returns to investors in the biggest technological transformation since the Internet era. In the analysts' view, the wave of AI computing power infrastructure construction and the strong profit growth driven by applications by the Magnificent Seven will continue to be an important force driving the overall EPS growth of the S&P 500 index. Additionally, according to senior market analysts at major Wall Street banks such as Morgan Stanley and Nomura, the main theme of the global AI super bull market, including US stocks, is evolving from "Nvidia's AI GPU demand frenzy outbreak" to "bullish wave for the entire AI computing power infrastructure chain." This will continue to support the US and global stock markets to revolve around the super bull market trend related to AI. The most typical example is undoubtedly Nvidia, whose latest performance clearly highlights that the global AI computing power infrastructure construction craze is far from over and is expanding from AI GPU/AI ASIC to data center CPUs, high-performance network infrastructure, enterprise-level HBM/DRAM/NAND storage, machine-level server clusters, AI super factories, and enterprise-scale large-scale AI cloud computing systems. On Wall Street, analysts' bullish sentiment towards the "global AI leader" Nvidia is growing stronger, with the average target price on Wall Street indicating that Nvidia's market value is expected to exceed $7 trillion. Therefore, the "cake" of the AI bull market is spreading beyond GPU core assets to the full-stack hardware bottleneck. Nvidia still sets the tone for the AI bull market, but PCBs, MLCCs, ABFs, ODMs, liquid cooling, power supplies, silicon wafers, CMP consumables, photoresists, glass substrates, SOI/InP, data center optical interconnect equipment, optical modules, and advanced packaging equipment may all undergo a revaluation. Following an 18% rebound from the low point triggered by the Iran conflict, the S&P 500 index is on track to achieve eight consecutive weeks of gains the longest streak since December 2023. However, there has been a slight rise in the US two-year Treasury yield, with Federal Reserve Governor Christopher Waller indicating that the possibility of the next Fed monetary policy rate action being an increase or decrease is equally likely. The currency market has fully priced in a hawkish expectation of a rate hike by the end of 2026. However, the price of US crude oil has experienced drastic fluctuations as traders try to determine when energy flow through the Strait of Hormuz will fully resume. After nearly three months of escalating Middle East geopolitical conflicts, Tehran is considering the latest proposal submitted by the United States. Pakistan has stated that its Chief of Army Staff is heading to the Iranian capital, indicating positive progress in the peace talks to end the war. Media reports suggest that a Qatari negotiating team has arrived in Tehran under US coordination to facilitate a long-term ceasefire agreement. US Secretary of State Marco Rubio has stated that there has been "some positive progress" in the negotiations with Iran, but "we are not there yet." The United Arab Emirates has been actively advocating for an end to the war, together with Saudi Arabia and Qatar, insisting on peaceful negotiations between the US and Iran rather than initiating a new round of conflict. According to senior analyst Craig Johnson at Piper Sandler, investors are actually overlooking macro headwinds and giving positive pricing and feedback surrounding peace prospects, which has been continuously providing tailwinds for the stock market. He said, "Global stock markets are showing a 'hope-driven' rebound." Mag 7 ignites profit engine! US stocks poised for the longest weekly consecutive gain since the end of 2023 With the momentum of the stock market shifting from AI computing power-driven gains to almost every sector of the US business community, the S&P 500 index is set for its strongest profit growth cycle in five years. According to Bloomberg Intelligence's latest statistics, about 93% of the companies in this benchmark index have reported earnings, with 83% of constituent companies significantly exceeding analysts' expectations. This is the highest level since 2021. Profit growth is broad-based, except for healthcare. Overall, the Mag 7 remains the main driver of profit expansion in the S&P 500 index, but the strong performance of the entire energy and technology sectors has overshadowed the sluggish consumer confidence related to the rise in oil prices driven by the war with Iran. Data compiled by Bloomberg Intelligence shows that the communication services and non-essential consumer goods sectors have brought the biggest positive surprises exceeding expectations, with materials and industrial sectors also outperforming. Senior analysts Nathaniel Welnhofer and Christopher Cain at Bloomberg Intelligence stated, "If cyclical sectors and non-AI sectors start contributing strong growth, while Nvidia and AI computing power infrastructure continue to generate profits, then 2026 may not look like a late-phase slowdown after a bull market cycle, but more like a replay of the super profit boom post the global COVID-19 pandemic in 2021." Looking at the whole year, profit growth in the S&P 500 index is expected to be concentrated on a few sectors revolving around tech stocks, while upward revisions in expectations for the energy, materials, and technology sectors are expected to drive the index to even stronger performance. As shown in the chart above, profit growth expectations for the full year in the S&P 500 index have been continuously raised - with the energy, technology, and materials sectors leading the overall earnings per share (EPS) growth trajectory of the index for 2026. Note: Data as of May 15. Bloomberg Intelligence stated that the majority of profit growth in the S&P 500 index is still coming from the likes of the Mag 7 big tech companies, and the AI computing power theme is seen as a major reason for the index's potential profit growth of over 20% in 2026. The S&P 500 index, with a weighting of approximately 40% for tech giants like Nvidia, Apple, and Microsoft, received significant index point and EPS upward revisions after the first quarter, with over 90% of constituent companies surpassing earnings expectations for four consecutive quarters. The performance and outlook of Nvidia have exceeded the consensus expectations on Wall Street, prompting analysts to continue to raise profit forecasts. The adjusted profit for this global market leader - which analysts currently expect to grow by about 84% this year, higher than the initial estimate of 64%. Citigroup raised its full-year profit expectations for Nvidia after the management projected sales could exceed $1 trillion by 2027 without including newer revenue sources such as the Groq LPX system and Vera processors. As shown in the chart above, Nvidia's adjusted profit forecast for the 2027 fiscal year has been raised. Note: Data as of May 22. According to the profit research director at LSEG, Tajinder Dhillon, the profit growth of the Magnificent Seven will continue to be an important force driving the overall EPS growth of the S&P 500 index, with expectations that they may stabilize their growth trajectory entering 2027, although this portfolio is still expected to significantly outperform the broader S&P 500 index. The chart below shows quarterly profit growth forecasts divided by portfolio - demonstrating the significant impact of the strong profit growth of the Magnificent Seven on the earnings fundamentals of the US stock market. This latest trend is strengthening the core view of bullish investors that the market will continue to focus on mega-cap tech giants, supported by fundamentals rather than on the eve of an "AI bubble" burst, and not driven by exuberance. Furthermore, due to the impact of the Iran conflict, the energy sector of US stocks has seen the largest upward revision. Analysts currently expect profit growth of 61% this year, higher than the initial estimate of 7.6%. Exxon Mobil and Chevron both reported stronger-than-expected first-quarter earnings, primarily due to higher oil and gas prices offsetting production interruptions related to the war. These, along with ConocoPhillips, are the major drivers of growth in this sector, accounting for about half of the sub-index weight. Brandon Bingham, an analyst at Wells Fargo, stated in a report that energy companies are generally optimistic about next year, citing strong project backlogs, stable growth, and further development potential as reasons for their positive outlook. Material stocks have become the third driving factor of the index, benefiting from price increases and supply tightening. Patrick Cunningham, an analyst at Citigroup, noted that disruptions in the Middle East have had limited impacts but have caused pressure on raw material costs. He also pointed out that demand trends appear stable, and companies maintain a cautious and constructive outlook. Cunningham mentioned that paint manufacturers Sherwin-Williams Co., PPG Industries Inc., and Axalta Coating Systems Ltd. all expect cost inflation, although they still anticipate moderate sales growth in the second half of the year. Chemical giant Dow Inc. remains conservative, but there is an upward trend if prices continue to improve. Bloomberg Intelligence stated that industrial gas producers such as Air Products and Chemicals Inc. and Linde Plc may get a boost, as a factory shutdown in Qatar in March led to a decline of about one-third of supply in the global helium market. The AI super bull market enters a "full-chain revaluation" moment: from Vera Rubin racks to glass substrates, GPUs no longer dominate the AI computing power theme! The AI computing power value chain beyond GPUs is taking off one after another As AI intelligent agents sweep across the globe, the investment theme of AI computing power is shifting from the "single-point computing race around AI GPUs" to the "full-stack computing system driven by AI intelligent agents." The next round of excess alpha returns will no longer be exclusive to the strongest players in the AI GPU/AI ASIC areas, but will systematically spread to data center CPUs, DRAM/NAND/HBM storage, AI PCBs, liquid cooling systems, data center optical interconnect systems, ABF substrates/glass substrates, and wide wafer foundry services across the full-stack AI computing power infrastructure layers. On April 30th, three cloud computing super giants, Microsoft, Google, and Amazon, all delivered outstanding results in a single night, highlighting the unexpected explosion of their cloud computing businesses fueled by the AI wave, prompting Wall Street to reevaluate the commercial returns of AI. The latest report from Morgan Stanley's analyst team suggests that the top five large-scale tech giants (Amazon, Google, Meta, Microsoft, Oracle) will have a combined capital expenditure of around $800 billion by 2026, which is expected to surpass $1.1 trillion in 2027, a revision from the previous estimate of $950 billion. The analysts at Morgan Stanley emphasized that behind these massive capital investments lies a core logic: first reinvest and build capacity, then rely on scaled commercial revenue based on AI computing resources and returns on invested capital; the surge in orders for cloud computing and AI application tokens is the most direct evidence that this logic is working. The rapid growth of their cloud computing businesses by these tech giants has led to Wall Street reevaluating the commercial returns of AI and the increasingly positive investment sentiment across the entire AI computing power industry chain. Wall Street's mainstream target prices for Nvidia have already priced it as a $7 trillion company; the most aggressive aggregate target prices push it to the $12 trillion level; and the core logic behind named high target prices is that Nvidia is upgrading from being the GPU leader to an "AI factory full-stack infrastructure platform" - GPUs, CPUs, networks, rack-level systems, software ecosystems, and capital returns collectively support valuation expansion, with the expectation that AI accelerated computing infrastructure spending may reach $3-4 trillion by 2030. Therefore, the main theme of the AI super bull market is shifting from the "Nvidia GPU frenzy outbreak" to the "AI infrastructure full-industry chain revaluation." In the previous phase, the market mainly bought into GPUs, HBM, cloud capital expenditures, and Nvidia's performance elasticity; but as per the Morgan Stanley research report, the tear-down analysis of the Rubin BOM tells investors that the value surge in the next-generation AI racks is not solely driven by GPUs, but by simultaneous increases in the value of PCBs, MLCCs, ABF substrates, power supplies, liquid cooling, ODM assembly testing, and other components. Morgan Stanley estimates that the Rubin VR200 NVL72 rack price is around $7.8 million, nearly twice the price of about $3.99 million for the GB300; in which the PCB content value has grown by 233%, MLCC by 182%, and ABF substrates by 82%. This signifies that the AI super bull market is entering the "component content expansion" second phase. From an engineering perspective, Rubin is not just a GPU replacement, but a leap in rack-level system complexity. When NVLink/NVSwitch, ConnectX, BlueField DPU, mid-boards, switch trays, harsh environments, and high-bandwidth memory and liquid-cooled power supplies enter the same rack, the value chain naturally shifts from GPU chips to high-speed copper connections, data center optical interconnect systems, power supplies, cooling, substrates, and high-layer PCBs. The proportion of GPU in the BOM actually decreases from around 65% for the GB200 to around 51% for the VR200, but the absolute amount for GPUs rises from around $2.52 million to around $3.96 million. This is the most crucial change in the AI industry chain: the leadership remains strong, but the "supporting links" are being revalued as indispensable performance bottlenecks. Nomura's research report reveals the structural changes at the deeper level of AI data centers: AI is shifting the semiconductor growth logic from "process miniaturization" to "structural innovation + material substitution + advanced packaging + optoelectronics." Previously, the capital market was more focused on advanced nodes at TSMC, Nvidia GPUs, and ASML lithography machines; but as we enter the era of 2nm, 1.4nm, and post-Moore's Law, backside power supplies, GAA/cFET, SoIC hybrid bonding, wafer-bonded NAND, glass core substrates, photonics on SOI, metal-oxide photoresists, InP/internal I/O silicon photonics, and other technologies are becoming key variables determining AI chip efficiency, bandwidth, yield, and cost. These explanations perfectly illustrate why the "AI super bull market" has not ended with the rise in Mag 7 valuations, but is evolving from a valuation narrative to profit realization and supply chain breadth expansion. Nvidia, Microsoft, Google, Meta, and Amazon continue to increase AI Capex, creating a systematic demand for AI server racks, advanced packaging, HBM, PCBs, MLCCs, ABF substrates/glass substrates, liquid cooling, power supplies, and data center power chains. Unlike in 2023-2024, the market is no longer just betting on "who has the strongest model/strongest GPU," but is reevaluating every link of the AI infrastructure that gets stuck due to power density, interconnect bandwidth, packaging yield, and supply bottlenecks. Glass substrates and data center optical interconnects are particularly worth mentioning. As the size of AI chip packages grows larger, the number of I/Os increases, and the signal rates rise, traditional organic ABF substrates and first-generation silicon interposers are under pressure due to warping, thermal expansion, signal loss, and large area flatness issues. Therefore, glass core substrates are seen as important candidates for next-generation advanced packaging. TrendForce points out that companies like TSMC, Samsung, Rapidus, and others are advancing glass interposer/glass substrate solutions, and SK Absolics aims to achieve mass production of glass substrates by 2026; IDTechEx indicates that AI/HPC demand is forcing the packaging stack to carry higher currents, more I/O, and higher signal rates, moving glass core substrates from niche technology to commercialization. Optical interconnects are a natural result of AI data centers transitioning from "compute stack" to "network bottleneck." After training and inference clusters scale to almost limitless sizes, the data flow between GPUs, racks, and data centers increases rapidly, while electronic interconnects face limitations in distance, power consumption, and bandwidth density. Technologies like optical interconnects (mainly including co-packaged optics, silicon photonics switches, optical circuit switches) and more advanced silicon photonics-level optical I/O technology can replace electrical signaling with optical signaling, significantly increasing bandwidth density and efficiency in large-scale AI training/inference networks, reducing latency and power consumption. This growing demand for higher optical interconnect capacities is what defines the common ground between GPU and TPU clusters. In other words, the next stage of the AI super bull market is not just about chips becoming faster, but about chips, advanced packaging systems, network infrastructure, optical interconnect systems, materials, liquid cooling, PCBs, MLCCs, ABF substrates/glass substrates, rack-level compute clusters, and power systems collectively determining the future of the "AI factory economics" initiated by Nvidia's CEO Jensen Huang.