The wave of AI computing power is undergoing a fundamental structural transformation. Investors are shifting their focus from "who is building the infrastructure" to "who is deploying AI." Based on Q2 2026 financial results, five representative AI-related stocks—NVIDIA, Broadcom, Marvell Technology, Dell Technologies, and Hewlett Packard Enterprise—are participating in this historic shift with distinct roles and growth trajectories. Their performance data not only reflect their own business outcomes but also serve as a mirror for the overall flow of AI infrastructure spending.
Comparison of financial data for five core AI-related stocks in 2026
NVIDIA: Growth Has Limits, but the Market Sets Higher Expectations
As the undisputed leader in AI computing power, NVIDIA delivered solid results in Q1 FY2026: revenue reached $44.1 billion, up 69% year-over-year, beating analysts’ expectations of $43.29 billion. The data center segment remains the main pillar, contributing $39.1 billion in revenue—an increase of 73% year-over-year. Notably, the Blackwell architecture chips are penetrating the market much faster than anticipated, accounting for 70% of data center revenue. The rapid transition from the H series to the B series has surprised the industry.
However, strong performance didn’t translate into a corresponding rise in share price—after the earnings release, NVIDIA’s stock fell about 1% in after-hours trading. The issue isn’t NVIDIA’s growth itself, but the slight gap between its growth rate and Wall Street’s "perfect expectations," which the market amplified. Revenue growth slowed from 78% in the previous quarter to 69%, attracting significant attention. Additionally, demand for H20 chips in China declined due to export controls, leading to a $5.5 billion asset impairment. This variable remains uncertain for the future.
NVIDIA is expanding its narrative. The Blackwell MVL72 AI supercomputer is now in production, and the Vera CPU opens new serviceable market space. Yet, as growth stocks are repriced for their growth rates, NVIDIA must demonstrate not just outperformance in the next quarter, but a clear growth path for the next two to three years. From a valuation perspective, with a current market cap of about $2.7 trillion, NVIDIA must continually prove it can capture market share across all layers of AI infrastructure spending—from GPUs to networking to CPUs.
Broadcom: Deepening the Moat of Custom Chips, but the Market Demands "Perfection"
If NVIDIA represents the general-purpose GPU route, Broadcom embodies another differentiated approach in AI chips—custom ASICs (application-specific integrated circuits). Q2 FY2026 results highlight the explosive power of this strategy: total revenue hit $22.19 billion, up 48% year-over-year and a record high; AI semiconductor revenue reached $10.8 billion, soaring 143% year-over-year.
Even more important is the visibility of its customer base. CEO Hock Tan confirmed Broadcom has six core custom chip customers, including Google, Meta, Anthropic, and OpenAI—tech giants with the most aggressive global AI infrastructure investments. These relationships aren’t short-term; they are deep, exclusive partnerships, with each XPU development cycle lasting 18 to 24 months and extremely high switching costs.
Despite this, Broadcom’s share price plunged over 13% after the earnings release. The market’s rejection stemmed from two minor misses: total revenue of $22.19 billion was slightly below Wall Street’s expectation of $22.27 billion, and infrastructure software revenue of $7.18 billion fell short of the $7.32 billion forecast. Furthermore, Hock Tan didn’t raise the full-year AI semiconductor revenue target above $100 billion, though buy-side institutions had already priced in an increase. The market is now evaluating custom chip players with the same rigorous standards as NVIDIA—even minor misses trigger rapid reassessment.
Marvell Technology: Valuation Pressure After Low-Base Growth and Long-Term Order Visibility
Marvell is the smallest among the five companies, but its growth story is the most resilient. FY2026 revenue reached $8.195 billion, up 42% year-over-year, with the data center segment accounting for 75%. The company has raised its FY2027 revenue guidance to nearly $11.5 billion, up $500 million from previous forecasts—about 40% annual growth.
Jensen Huang publicly called Marvell the "next trillion-dollar company" at Taipei Computex, sparking a single-day stock surge of over 32% in early June 2026. Year-to-date gains approached 239%. However, the stock soon saw a 15% single-day pullback amid wide fluctuations. Valuation battles are especially intense at this stage, with Marvell’s current P/E ratio exceeding 70x and peer valuations being reassessed.
Marvell and Broadcom have significant business overlap, both focusing on providing ASIC design services and optical interconnect solutions for hyperscale cloud providers. The difference lies in Broadcom’s larger scale and more stable cash flow, while Marvell offers greater growth flexibility, with its AI segment still poised for further expansion. Marvell’s valuation re-rating in the ASIC space depends on whether AWS and Google continue to raise capital expenditure guidance in upcoming earnings reports—a key indicator for sustained high AI inference demand.
Dell Technologies: The Biggest Cyclical Beneficiary of AI Server Infrastructure
Dell is the fastest-growing among the five, with standout recent financial results. For the quarter ended May 1, revenue reached $43.84 billion, up 88% year-over-year—the largest single-quarter growth since the company returned to public markets. Adjusted earnings per share came in at $4.86, far exceeding the $2.94 consensus estimate.
AI server business is the core growth engine: quarterly revenue soared to $16.1 billion, up 757% year-over-year. The full-year outlook for AI-optimized server revenue was raised from $50 billion to $60 billion. After the earnings release, Dell’s stock jumped nearly 33%, driving simultaneous surges for HPE, Super Micro, and other server manufacturers.
Dell’s growth is driven by the shift in global enterprise AI deployment from "pilot phase" to "scaled deployment." Vice Chairman and COO Jeff Clarke noted on the earnings call that traditional server shipments also saw significant growth, as semiconductor companies and tech giants leverage servers to support internal inference and agent workloads. Dell’s order backlog has climbed to $51.3 billion, proving the sustainability of infrastructure spending rather than a one-off surge.
A constraint to consider is the gross margin structure. The rising proportion of low-margin AI servers pushed gross margin down to 17.7%. As AI infrastructure investment continues to expand, Dell’s main logic is market share growth and revenue scale expansion, but cost structure pressure remains an unavoidable constraint in its valuation thesis.
HPE: Early Validator of Enterprise AI Infrastructure Procurement Cycles
HPE’s latest earnings report provides the clearest signal that enterprise AI procurement has entered the scaled deployment phase. Q2 FY2026 revenue rose 40% year-over-year to $10.7 billion, beating the $9.79 billion consensus. Adjusted earnings per share were $0.79, nearly 50% above both its own guidance range and market expectations.
The biggest information gain comes from AI order backlog structure: HPE secured $1.8 billion in new AI system orders, bringing total AI system orders to $16.4 billion and pushing overall backlog to a record $5.9 billion. The company also raised its full-year revenue growth guidance from 17%–22% to 29%–33%. CEO Antonio Neri stated the company’s performance has outpaced its long-term financial plan by a full two years.
The integration of Juniper is a key recent variable for HPE. The $14 billion acquisition transformed HPE from a server-focused hardware manufacturer into an integrated infrastructure provider with a networking profit engine. Networking revenue reached about $2.7 billion, up 152% year-over-year, with an operating margin of 23.7%—nearly double the company’s overall operating margin. With Juniper’s integration, HPE is now positioned to compete with Cisco in the AI-native networking space, addressing network bottlenecks as AI cluster sizes expand.
Gate Real Stock Trading: A Convenient Channel for Allocating AI-Related Stocks
For investors, analyzing AI-related stock data is foundational, but participating in trading efficiently and cost-effectively is the key to execution. Gate now offers real stock trading, allowing users to trade over 10,000 stocks and ETFs from five major exchanges, including NYSE and NASDAQ, directly with USDT.
Unlike traditional CFD or contract products, Gate’s stock trading involves actual stock assets. Trades and clearing are handled through partner broker Alpaca, so buying means holding real stock ownership and receiving dividends and shareholder rights. There are no funding rates or overnight holding fees. For investors seeking long-term allocation of AI-related stocks, this eliminates the rolling costs or swap fees associated with contract products. You can switch from USDT to US stock assets within the same crypto account, lowering the operational barriers of opening traditional securities accounts and transferring funds across markets. The fractional share trading feature, starting from as little as 0.01 shares, allows investors to benefit from the growth of high-priced stocks like NVIDIA without purchasing whole shares.
Conclusion
Today’s AI-related stock trading is shaped by two intertwined forces: first, fundamental growth driven by each company’s products and technology strategies; second, investors’ collective pricing of the long-term visibility of AI infrastructure spending. AI infrastructure investment remains in a relatively early stage, and trillions of dollars are expected to flow into this sector over the coming years—even though this spending cycle has lasted about two and a half years, both consumer and enterprise AI adoption are still accelerating, especially as inference demand surges with larger and more complex models.
Server spending is projected to grow 36.9% in 2026, and the boundaries of AI-related stock growth continue to expand. The market’s pricing logic is evolving: investors are moving from broad AI themes to a stage of fine-tuned selection between winners and laggards. It’s important to note that investing in AI-related stocks involves high industry concentration, rapid technological iteration, and risks related to policy and export controls. Past performance does not directly indicate future returns. This article provides objective industry and data analysis only and does not constitute any trading advice. Investors should make independent decisions based on their own risk tolerance.
Both crypto assets and stock trading involve market volatility risks. Gate’s stock trading services are cleared through partner brokers, and user assets are protected under the relevant regulatory framework. Please refer to official announcements for specific trading rules and fee schedules.
