Between May and June 2026, the global semiconductor industry underwent a dramatic valuation reset. NVIDIA and Broadcom—the two most representative companies in the AI chip sector—released their earnings reports less than three weeks apart, yet received sharply contrasting reactions from the market.
NVIDIA reported FY27 Q1 revenue of $81.6 billion, up 85% year-over-year, with data center revenue surging 92% year-over-year. Its Q2 revenue guidance of $91 billion exceeded market expectations. Although the stock price experienced brief volatility after the report, Wall Street’s 62 analysts still set an average target price as high as $298.
Broadcom posted FY2026 Q2 revenue of $22.19 billion, up 48% year-over-year, with AI semiconductor revenue reaching $10.8 billion—a staggering 143% increase year-over-year. However, the company’s Q3 AI semiconductor revenue guidance came in at $16 billion, falling short of analysts’ $17.2 billion consensus. This "miss" directly triggered a 10.3% single-day plunge in the PHLX Semiconductor Index on June 5, wiping out over $1.3 trillion in market capitalization from the sector.
Both companies delivered AI-driven growth, so why did the market reward NVIDIA’s "beat" but punish Broadcom’s "high growth but slightly below expectations"? The core variable here isn’t the absolute numbers, but rather the difference in expectation management.
Two Earnings Reports, Two Fates
NVIDIA: Beating Expectations Becomes the New Normal
NVIDIA’s FY27 Q1 numbers set records across nearly every metric. Revenue reached $81.615 billion, up 85% year-over-year and 20% quarter-over-quarter. Data center revenue totaled $75.2 billion, accounting for over 92% of total revenue, up 92% year-over-year and 21% quarter-over-quarter.
Structural changes are even more noteworthy. Data center compute revenue was $60.4 billion, up 77% year-over-year; data center networking revenue hit $14.8 billion, up 199% year-over-year, both all-time highs. This indicates that AI demand has expanded from pure GPU computing power to encompass the entire AI infrastructure stack—networking, rack-scale systems, optical communications, power, and cooling are now all integral to AI factories.
For FY27 Q2, NVIDIA guided revenue of $91 billion (±2%), well above the pre-earnings market consensus of $86–87 billion. Notably, the company specified that this guidance excludes data center compute revenue from China. This detail is crucial—being able to provide such robust guidance even without China shows that AI compute demand in other regions is strong enough to sustain high growth on its own.
Broadcom: Impressive Numbers, Even Higher Expectations
By any objective measure, Broadcom’s earnings were outstanding. Q2 total revenue reached $22.19 billion, up 48% year-over-year; semiconductor revenue hit $15 billion, with AI revenue at $10.8 billion—up 143% year-over-year and 28% quarter-over-quarter. On the earnings call, CEO Hock Tan described AI semiconductor demand as "Simply insatiable."
Broadcom received over $30 billion in AI semiconductor orders in Q2, but shipped only $10.8 billion during the same period. The company expects full-year FY2026 AI semiconductor revenue to reach $56 billion, up about 180% from FY2025; for FY2027, AI semiconductor revenue is projected to exceed $100 billion.
However, the market’s main focus was on Q3 AI semiconductor revenue guidance—$16 billion, below analysts’ $17.2 billion consensus. That $1.2 billion gap set off a chain reaction.
The Divergent Logic of Expectation Management
The difference in market response to NVIDIA and Broadcom boils down to their philosophies on expectation management.
NVIDIA’s strategy is to "guide expectations, then deliver a beat." Even before the earnings release, market consensus for NVIDIA’s revenue was already high—around $86.5 billion. NVIDIA responded with actual results of $81.6 billion and guidance of $91 billion, not only beating expectations for the current quarter but also signaling continued growth momentum with guidance far above market forecasts.
Broadcom’s approach was to "deliver high growth, but fall short of the latest market expectations." In the five trading days leading up to the earnings release, Broadcom’s stock had already risen on the back of lofty AI business expectations. When the Q3 AI guidance of $16 billion was announced—7% below the $17.2 billion consensus—even this small gap was magnified in an AI chip sector where "beating expectations" has become the norm.
A deeper difference lies in the predictability of each company’s business. NVIDIA’s general-purpose GPUs serve a broad range of AI training and inference scenarios, with customers spanning nearly all cloud providers and enterprise users. This results in highly visible and diversified demand. Broadcom’s custom ASIC business, on the other hand, is heavily reliant on a handful of hyperscale customers—Google, Meta, and ByteDance are key drivers. High customer concentration means greater revenue volatility, and the market demands more precise guidance.
The $1.3 Trillion Domino Effect
Broadcom’s earnings miss triggered a swift sell-off across the semiconductor sector. On June 5, the PHLX Semiconductor Index (SOX) plunged 10.3%—the largest single-day drop since the pandemic-induced panic in March 2020.
Broadcom’s own shares fell about 14%, erasing roughly $286 billion in market value. NVIDIA dropped around 7%, slipping below the $5 trillion market cap threshold. Micron Technology fell about 13%, AMD about 10%, and Intel about 11%. Over two trading days, the PHLX Index lost a cumulative 12%.
While Broadcom’s guidance miss was the catalyst, deeper factors were at play. The semiconductor sector had experienced explosive gains for months, with the PHLX Index up more than 300% since late 2022. Valuations were at historic highs—the VanEck Semiconductor ETF was trading at a 67.2% premium to its intrinsic value. At the same time, strong US employment data fueled concerns that the Federal Reserve would keep interest rates elevated, putting additional pressure on high-valuation tech stocks.
This was a classic case of "high expectations + high valuations + minor disappointment" triggering a systemic pullback. Capital rotated out of AI semiconductor stocks and into traditional sectors like banking and industrials, with the Dow Jones Industrial Average hitting a record high the same day.
Structural Shifts in the AI Chip Competitive Landscape
The Broadcom-triggered sell-off also reflects deeper changes underway in the AI chip competitive landscape.
NVIDIA currently holds about 75–80% market share in AI chips—down from 87% in 2024, but still overwhelmingly dominant. Morgan Stanley analysts note that NVIDIA’s market share "hasn’t budged" in the past two years, cementing its industry leadership. The CUDA ecosystem remains a formidable moat that competitors struggle to breach.
Broadcom, meanwhile, has pursued a custom XPU/ASIC chip strategy, focusing on meeting the specific needs of top-tier customers. Google’s TPU, Meta’s MTIA, and OpenAI’s in-house chip initiatives are all major orders for Broadcom. The advantage of this model lies in high customer stickiness and attractive margins; the downside is customer concentration and greater revenue volatility.
AMD directly competes with NVIDIA in general-purpose GPUs, with Q1 2026 data center revenue reaching $5.8 billion, up 57% year-over-year. However, it still trails NVIDIA by a significant margin in overall market share.
Morgan Stanley forecasts that by the end of 2026, inference workloads could account for 50–60% of AI chip demand, with NVIDIA’s share of the inference market potentially dropping from 70% in 2025 to 60–65%. The rise of inference will create more growth opportunities for custom chipmakers like Broadcom, but in the near term, NVIDIA’s dominance in training remains unshaken.
Gate Stock Trading: A New Gateway to AI Sector Investment Opportunities
For users looking to invest in the AI chip sector, Gate’s stock trading service offers a low-barrier, highly efficient entry point.
Coverage and Trading: Gate’s real-stock trading covers over 11,500 stocks and ETFs, including AI chip leaders like NVIDIA (NVDA), Broadcom (AVGO), AMD, Micron Technology, and a variety of sector ETFs. Users can buy and sell directly from the familiar crypto trading interface—no platform switching required.
Key Advantages:
- Ultra-low fractional share threshold: Start investing in US stocks with as little as 0.01 shares, for just $1.
- USDT settlement: Trade US stocks directly with USDT, eliminating the hassle of "selling crypto → withdrawing fiat → cross-border remittance → broker funding."
- Regulatory compliance: All stock trades are executed by Alpaca, a fully licensed US Broker-Dealer and clearing firm, with full SIPC protection.
- Unified account management: Manage both crypto and stocks in a single account for more efficient asset allocation.
How to Trade: After logging into Gate, select the "Stocks" trading pair in the trading interface, enter the stock ticker (e.g., NVDA, AVGO), and buy or sell at real-time market prices using USDT. The trading process closely mirrors crypto trading, significantly lowering the learning curve for traditional US stock investing.
Conclusion
The divergent market fates of NVIDIA and Broadcom between May and June 2026 exemplify the AI chip sector’s entry into a "sensitivity to expectation management" phase. When sector valuations have fully priced in the "limitless AI demand" narrative, the market reacts disproportionately to any signal that falls short of expectations.
NVIDIA has proven the sustainability of its growth momentum with both a strong earnings beat and even stronger forward guidance. Broadcom, despite delivering a record-breaking quarter, saw a minor shortfall in Q3 AI guidance trigger a systemic sell-off at lofty valuations. This isn’t a reversal of fundamentals, but rather a recalibration by the market under extreme pricing.
From a longer-term perspective, the AI chip growth story is far from over. Broadcom CEO Hock Tan reiterated on the earnings call that FY2027 AI semiconductor revenue will exceed $100 billion; NVIDIA, with its full-stack capabilities across GPU, networking, and CPU, continues to lead the field. The two companies represent distinct technological paths and business models in the AI chip sector—general-purpose GPUs versus custom ASICs—and will continue to coexist and compete for years to come.
For investors, short-term volatility in AI chip stocks shouldn’t obscure the long-term structural growth story. On Gate, users can flexibly allocate to core AI chip assets and seize opportunities in this supercycle through low-barrier fractional trading and direct USDT settlement.
