Amid US-Iran tensions, AI investment continues to accelerate.The four US Big Tech companies (Microsoft, Amazon, Google, Meta) raised their forecasted capital expfinishiture for 2026 in their earnings announcement after the close on April 29.
However, market evaluations were mixed. $Alphabet-C (GOOG.US)$ rose about 7% in after-hours trading, while $Meta Platforms (META.US)$ fell by about 7%, clearly dividing stock prices within the same expansion phase. The essence of this difference lies in how much AI investment exceeded market expectations in terms of performance.
Furthermore, when each company examinesthe factors indicating an expansion in capital expfinishiture,toUpon closer examination, the modifying nature of this investment becomes clear. In addition to growth investments to meet expanding AI demand, addressing supply constraints in areas such as memory, power, and data centers has become unavoidable.Capital investment is transitioning to a dual-layer structure of ‘growth investment’ and ‘response to supply constraints’.This represents a significant shift indicating that the essence of AI investment itself is modifying, rather than just an expansion in investment.
This article will explore,The latest developments surrounding the earnings reports and AI investments of the Big Tech four companieswill be reviewed, whilebeneficiary stocks and impacted stockswill be organized.
Big Tech four companies’ performance and immediate stock price reactions
– Earnings and outsee that split market expectations
The US Big Tech four companies have collectively raised their capital expfinishitures, signaling intentions to further expand AI-related investments, but stock price reactions were clearly divided. $Alphabet-C (GOOG.US)$ rose approximately 7% in after-hours trading, while $Meta Platforms (META.US)$ It fell by about 7%. The key point of evaluation lies not in the scale of investment, but in how much earnings have responded to market expectations.
$Alphabet-C (GOOG.US)$Cloud revenue reached $20 billion, a year-over-year increase of63%, surpassing market expectations (50% increasefar exceeded expectations. Moreover, the order backlog has sharply expanded to $460 billion, indicatingincreased visibility of demandis progressing. During the earnings briefing, the CFO stated that ‘AI computing demand is at an unprecedented level both inside and outside the company,’clear alignment between investment and demandis evident.
$Amazon (AMZN.US)$ AWS revenue grew by a year-over-year28% increase, reaching $37.6 billion, surpassing market expectations (26% growth). Additionally, contracts with OpenAI and Anthropic are expected to result in over $100 billion in spfinishing in the coming years. ManagementAI demand is driving cloud growthand the possibility of investment recovery has been somewhat demonstrated.
$Microsoft (MSFT.US)$ Azure’s revenue grew approximately39% year-over-year, slightly surpassing market expectations ofa 38% increase. AI-related sales have expanded to an annualized scale of $37 billion, but during the earnings call, the CFO stated that ‘capacity constraints are expected to continue through 2026,’ highlighting concerns about not fully capturing demand. As a result, the stock price finished without clear direction.
$Meta Platforms (META.US)$ Revenue reached $56.3 billion, surpassing market expectations, but raising the capital expfinishiture forecast to $125-$145 billion was perceived negatively. During the earnings call, management explained that ‘rising component prices and increasing data center costs were contributing factors.’ In addition,there was limited detail provided on monetizing AI investments,and shares fell by about 7% after-hours.
The market is evaluating the Big Tech four not just on the scale of their capital expfinishitures, but also on outperformance relative to market expectations, forecasts, and the linkage between investment and profitability.
The reality of expanding capital expfinishitures
AI investment reaches its largest-ever scale
Capital expfinishitures by the four Big Tech companies leading AI investments are projected to reach historic levels by 2026.
$Microsoft (MSFT.US)$ has been increasing its capital expfinishiture quarter by quarter, and for 2026 it expects approximately USD 190 billion in capital expfinishitures. $Amazon (AMZN.US)$ is advancing investments centered around data centers with a scale of around USD 200 billion, displaying a significant increase from the previous year due to growing AI demand. $Alphabet-C (GOOG.US)$ revised its capital expfinishiture forecast from the previous USD 175 billion to 185 billion to USD 180 billion to 190 billionand are strengthening their response to AI demand. $Meta Platforms (META.US)$also$115 billion to $135 billion to $125 billion to $145 billionupwardly revised.
What is the essence of this expansion?
Examining the content of each company’s earnings briefing regarding the factors behind these upward revisions reveals a clear difference from the past. Namely,this investment expansionis not only the conventional ‘forward-seeing investment in growth expectations’ but alsoan ‘investment for growth backed by already manifest demand’andSimultaneouslyan ‘investment aimed at addressing supply constraints.’This is where it lies.
AI demand is rapidly expanding.
$Alphabet-C (GOOG.US)$ explained that AI computing demand has reached an all-time high both internally and externally, with cloud revenue growing 63% year-over-year to $20 billion and the order backlog sharply expanding to $460 billion. The expansion of AI services centered around Gemini is also driving growth.
$Amazon (AMZN.US)$ AI demand is also at the core of cloud growth. AWS revenue exceeded market expectations with a 28% year-over-year increase to $37.6 billion, driven by expanding demand from AI startups and large enterprises. Long-term contracts with OpenAI and Anthropic are expected to generate $100 billion in demand over the next few years.
$Microsoft (MSFT.US)$ AI demand remains strong, with Azure growing by about 40%, and AI-related revenue reaching an annualized scale of $37 billion. However, constraints on data center capacity persist, preventing full capture of demand.
$Meta Platforms (META.US)$ Demand is also expanding through the utilize of AI for advanced ad delivery and recommfinishations, but much of this remains within its own services and is not yet visible as revenue like it is for cloud services.
Supply-side constraints are also a key factor driving investment expansion.
Supply-side constraints have also become a significant factor in investment expansion.
$Microsoft (MSFT.US)$ acknowledged a cost increase of approximately $25 billion due to rising component prices, $Meta Platforms (META.US)$ similarly identified rising component prices as a factor behind the expansion of capital expfinishitures. Additionally, Microsoft pointed out that a shortage of data center capacity is constraining growth, while Meta is also expanding investments due to server supply tightness and increasing infrastructure loads.
$Alphabet-C (GOOG.US)$ is also aware of the impact of supply constraints. In earnings briefings, they mentioned operating in a complex supply chain environment and explained that depreciation expenses, data center operating costs, particularly electricity costs, are pressuring profits as AI infrastructure investments grow. They also indicated a rising proportion of investment in servers and network equipment, highlighting infrastructure acquisition as a critical challenge.
$Amazon (AMZN.US)$ Similarly, investment is proceeding based on the assumption of constraints on the supply side. In the earnings briefing, they emphasized a strategy to reduce depfinishence on external GPUs by introducing Trainium, a custom chip, aiming for cost containment and supply stability. This is not merely an efficiency improvement but a structural response to supply constraints.
In this way,Capital expfinishiture is being driven not only as a response to demand expansion, but also as a measure to address rising costs and supply constraints.。AI investment has entered a phase where ‘growth investment’ and ‘constraint-response investment’ coexist,and its nature is clearly modifying.The US Big Tech four are expected to continue expanding their capital investments into 2027,andthe current structure may continue into next year.。
Beneficiary and impacted stocks in the new ‘AI investment’
A two-layer structure of ‘growth investment + constraint response’ generating capital flows
Based on the recent earnings reports of the four Big Tech companies, the shift in AI investment has become apparent,Beneficiaries and impacted stocks of AI investmentare expected to display Need to be organized from the perspective of a two-layer structure: ‘growth investment’ and ‘constraint-response investment’.It seems necessary.
“Growth Investment”refers to areas that are expanding through the growing utilize of AI, cloud computing, and advancements in AI models, driven by an increase in demand.
“Constraint-Driven Investment”addresses supply-side bottlenecks such as memory shortages, power and cooling issues, and data center capacity limitations.
The current AI investment trfinish displays that both aspects are progressing simultaneously. The benefits are not limited to simple demand expansion but are also spreading into areas central to addressing supply constraints. Therefore, it is important to note that investment opportunities are distributed not only among ‘AI growth stocks’ but also among ‘infrastructure and resource stocks supporting AI,’ which have already been accelerating since the launchning of the year.
Beneficiary Stocks – Recipients of AI Growth Investment
In the semiconductor field, first of all,GPUs continue to play a central role, $NVIDIA (NVDA.US)$ GPUs maintain a dominant presence in AI training and some inference applications, remaining at the core of growth fueled by increasing demand from data centers.
Second,as inference processing expands, the importance of custom chips (ASICs)is rapidly increasing. $Broadcom (AVGO.US)$ and $Marvell Technology (MRVL.US)$ is becoming an indispensable presence in custom AI accelerators for hyperscalers and high-speed connectivity chips within data centers.
Moreover,the importance of CPUs is also being reassessed. $Amazon (AMZN.US)$During the earnings briefing, mentioned that computational loads are being distributed from GPUs due to the rise of agent-based workloads, real-time inference, and complex processing, leading to an expanded role for CPUs. As a result,$Advanced Micro Devices (AMD.US)$and$Intel (INTC.US)$ is being reevaluated as a key player supporting part of the AI workload through its data center-tarobtained CPUs.
Furthermore,$Arm Holdings (ARM.US)$Many custom chips adopt as their design foundation, and from the perspective of design licensing, structural benefits are expected.
Foundry and manufacturing equipment
On the manufacturing side $Taiwan Semiconductor (TSM.US)$occupies the most critical position. Many AI semiconductors rely on leading-edge processes, and the company’s advanced logic manufacturing capabilities are at the heart of supply constraints. Demand for advanced nodes is surging due to AI applications, creating a structural advantage that benefits both utilization rates and pricing.
Semiconductor manufacturing equipmentManufacturers are also a structurally advantageous area. $ASML Holding (ASML.US)$ holds a monopolistic position in EUV lithography systems and is essential for the expansion of advanced processes.$Applied Materials (AMAT.US)$and$Lam Research (LRCX.US)$ It plays a crucial role in the film formation and etching processes, and demand is expanding due to the increased production of advanced semiconductors for AI. Furthermore, $KLA Corp (KLAC.US)$ In the inspection and measurement processes, it is an indispensable presence. As process complexity increases, yield management becomes more critical, leading to a structural expansion in demand for the company’s equipment.
In this way,Benefits of AI growth investmentisIn addition to each layer of semiconductors such as GPUs, ASICs, and CPUs, the benefits extfinish across the entire supply chain from design and manufacturing to equipment.Particularly important is that with the expansion of inference and operational stages, semiconductor composition is becoming increasingly multi-layered.
Beneficiary Stocks – Group of stocks facing supply constraints
Memory and storage-related
Memory sectoris the area benefiting the most as a core of supply constraints. $Amazon (AMZN.US)$ During the earnings briefing, the rapid expansion of AI workloads has led to a tightening of memory supplies such as DRAM and NAND,and its impact is expected to continue at least until around2027.has expressed this view.
announced its earnings on April 29th $Seagate Technology (STX.US)$ reflecting these modifys in supply and demand structure,emphasized the strength of storage demand for data centers.Especially due to the explosive increase in data volume for AI applications, demand for high-capacity HDDs is rapidly expanding,Increased shipments and price improvements of related productsare boosting profits. Additionally, sustained strong demand from cloud and hyperscalers indicates that storage investments accompanying AI infrastructure expansion will be a mid-term growth driver.
In such an environment, $Micron Technology (MU.US)$ companies like Samsung Electronics/SK Hynix (related ETF: $iShares MSCI South Korea ETF (EWY.US)$ ）、 $SanDisk (SNDK.US)$ 、 $Western Digital (WDC.US)$ are directly impacted by the surge in demand from AI data centers. AI servers are absorbing a large portion of DRAM and NAND supplies, leading to price increases that directly translate into profit growth.
Power and cooling (physical constraints of AI infrastructure)
Electric power and cooling sectorsare equally important. $Vertiv Holdings (VRT.US)$and$Eaton (ETN.US)$ is expected to see structural growth due to increased power and cooling demands stemming from the expansion of data centers. In the expansion of AI infrastructure, power and cooling have become clear bottlenecks, and investments in this area are likely to be long-term.
Data Center Network (Response to Explosive Traffic Growth)
Facilities and network fields related to data centersWell then,$Arista Networks (ANET.US)$and$Cisco (CSCO.US)$ will benefit. The increase in AI workloads has led to a surge in data traffic, creating the advancement and expansion of network infrastructure essential.
In addition, the shortage of data center capacity itself is becoming apparent. For this reason,Supply of AI-dedicated data centersis also benefiting the players involved. $NEBIUS (NBIS.US)$and$CoreWeave (CRWV.US)$ Companies that provide infrastructure specialized for AI workloads are rapidly capturing demand and emerging as complementary forces to hyperscalers. The key point here is that the shortage of data centers is not just a constraint but is creating new growth areas.
Next-Generation Energy (Securing Power Supply)
From the perspective of bottleneck resolution,Areas related to securing power supply itselfare seeing an expansion in investment tarobtains. $Alphabet-C (GOOG.US)$ The fact that moomoo highlighted the rise in power costs during its earnings briefing indicates that ensuring the availability of power sources is becoming an important theme, shifting beyond mere efficiency improvements.
$Consnotifyation Energy (CEG.US)$ is gaining attention as a stable power supply source for data centers, while in the compact modular reactor (SMR) sector, $Oklo Inc (OKLO.US)$and$NuScale Power (SMR.US)$ are being anticipated as dedicated power sources for AI data centers. Additionally, amid the trfinish of returning to nuclear power, $Uranium Energy (UEC.US)$and$Centrus Energy (LEU.US)$ fuel supply-related companies such as are also expected to benefit in the medium to long term from the perspective of easing supply constraints.
Optical interconnect (the next communication bottleneck)
Internal data center communicationsare seeing a rapid increase in the importance of optical connections as a bottleneck. As communication volumes explode due to the rise in AI workloads, the shift from traditional copper wiring to fiber optics is progressing, and there is growing demand for next-generation standards such as 1.6T transceivers.
$Lumentum (LITE.US)$and$Coherent (COHR.US)$ holds high competitiveness in optical components and laser technology, and $Credo Technology (CRDO.US)$ is well-positioned to capture the growth in this area by providing chips for high-speed data connectivity.
Affected stocks – The ‘victims’ of supply constraints
Impact of supply constraintsis particularly evident in non-AI fields. $Dell Technologies (DELL.US)$and$HP Inc (HPQ.US)$ PC-related companies such as are being affected by memory shortages and rising component prices, with some products seeing price hikes of up to 30%. As AI servers are prioritized, supply for non-AI applications is being pushed back, leading to tightening supply-demand conditions and simultaneous cost increases.
$Apple (AAPL.US)$ may similarly be affected by memory (DRAM/NAND) shortages and rising component costs. Particularly, as memory supplies concentrate on AI data centers, pressure on product costs is expected to increase. However, in the medium to long term, reduced infrastructure depfinishency through self-developed ‘Apple Silicon,’ combined with added value from integrating generative AI capabilities into devices, could create opportunities to capture demand by accelerating replacement cycles. In the earnings report after the close on April 30 local time, the balance between the impact of rising costs and the contribution of AI features to sales will likely be a key focus.
Additionally, a similar impact is being observed in the non-AI server sector. The concentration of resources towards AI infrastructure has led to delays in the supply of traditional servers, which may result in postponements for certain parts of corporate IT investments.
Prepared by Market Analyst Julie on April 30, 2026
Source: Created by moomoo Securities based on materials from various companies, Bloomberg, and other sources
Leave a Reply