The top stock by trading volume on Friday was Micron Technology (MU), which closed down 13.21%, with a turnover of $67.9 billion. Memory chip stocks were broadly lower on Friday, with chipmakers declining after Broadcom's (AVGO) AI chip sales outlook failed to meet elevated market expectations.

Saxo's chief investment strategist Charu Chanana stated, "Broadcom is a trigger, reminding the market of how excessive expectations have become. Investors have priced in a lot of perfection for AI, so even a small disappointment can lead to a fairly sharp re-rating."

Additionally, a report released Friday by research firm SemiAnalysis suggested that the CPU-side system memory capacity for NVIDIA's (NVDA) next-generation Vera Rubin server platform would be nearly halved, sparking concerns about a slowdown in memory demand and leading to a collective pullback in memory-related stocks.

The second-ranked stock by volume was NVIDIA (NVDA), which closed down 6.20%, with a turnover of $44.965 billion. NVIDIA CEO Jensen Huang stated that the three major memory chip manufacturers have been certified to supply the most advanced high-bandwidth memory chips for NVIDIA's AI performance-boosting products. This decision means that SK Hynix, Samsung Electronics, and Micron Technology (MU) are poised to begin mass production and supply of HBM4 chips. These three companies dominate the global memory chip market and have been fiercely competing for share in this business. Huang, speaking to reporters upon arrival in Seoul for a multi-day visit, said all three suppliers have been certified and are already in production, racing to support NVIDIA's latest AI chip, Vera Rubin.

The third-ranked stock was Marvell Technology (MRVL), which closed down 16.74%, with a turnover of $24.865 billion. Marvell's shares fell sharply on Friday after a staggering rally of over 50% in just six trading sessions that took the stock to a record high, as investors opted to take profits following the recent massive gains.

The direct catalyst for Marvell's surge was NVIDIA CEO Jensen Huang. On June 2, at the Taipei International Computer Show, Huang publicly stated that Marvell has the potential to be "the next trillion-dollar company." Stimulated by these remarks, Marvell's stock soared over 32% that day, marking its largest single-day gain in history and adding over $50 billion in market value. Huang also emphasized that in AI data centers, connectivity is crucial, and Marvell's networking and connectivity chips play a key role. Just three months ago, NVIDIA announced a $2 billion investment in Marvell to deepen collaboration in custom AI chips and NVLink fusion technology.

The fourth-ranked stock was Tesla Motors (TSLA), which closed down 6.56%, with a turnover of $24.713 billion. A report on Friday indicated Tesla had postponed the live demonstration of its Roadster sports car to August.

JPMorgan upgraded Tesla's rating on Friday from "Underweight" to "Neutral" and significantly raised its price target from $145 to $475, more than tripling it. This major shift marks a fundamental adjustment in the Wall Street firm's valuation logic for Tesla—moving from traditional automotive business profitability to the long-term growth potential of autonomous driving and robotics.

The upgrade was led by analyst Rajat Gupta, who recently took over coverage of Tesla. JPMorgan's report noted that investors are looking beyond Tesla's slowing core EV business to focus on future growth engines like robotaxis, humanoid robots, AI chips, and software services, which could reshape the company's profit structure over the next decade.

Gupta deconstructed Tesla's valuation into five interconnected markets: automotive, energy storage, robotaxis, humanoid robots, and infrastructure licensing, estimating their combined total addressable market could reach around $3.9 trillion by 2035. The report emphasized that Tesla's unparalleled vertical integration capabilities in both hardware and software are its core competitive advantage, but the market still underestimates and misunderstands this to some extent.

The fifth-ranked stock was Advanced Micro Devices (AMD), which closed down 10.86%, with a turnover of $22.152 billion. High-profile tech stocks were generally under pressure on Friday.

On the news front, strong employment data released officially on Friday in the US sharply increased market bets on Federal Reserve rate hikes—traders have moved forward the timing of the first rate cut from a previous expectation of March next year to January next year. Additionally, escalating geopolitical tensions in the Middle East—Iranian naval forces conducted warning shots at a US warship in the Gulf of Oman—further dampened market risk appetite.

The sixth-ranked stock was SanDisk, which closed down 11.39%, with a turnover of $20.857 billion. The memory sector continued its decline from the previous day, falling across the board on Friday. On June 4, after Broadcom (AVGO) reported earnings but failed to raise its 2026 fiscal year AI chip sales guidance, it dashed the high market expectations that had pushed chip stocks to elevated levels. Broadcom plunged over 12% that day, and the selling sentiment quickly spread across the entire semiconductor sector, with memory-related stocks being the first to be hit.

Furthermore, two consecutive days of selling indicate that, with AI-related stock valuations at high levels, market sensitivity to AI-related performance guidance has increased significantly. Any signal that fails to exceed expectations can trigger profit-taking. In terms of sector rotation, the Dow Jones Industrial Average rose 1.73% against the trend on June 4 to a record high, with previously lagging sectors like healthcare and financials becoming "safe havens" for funds, further diverting liquidity away from tech stocks.

The eighth-ranked stock was Broadcom (AVGO), which closed down 7.92%, with a turnover of $19.912 billion. Broadcom has declined for three consecutive trading sessions after its recent AI chip sales guidance fell short of market expectations, triggering a collective sell-off in the semiconductor sector.

Broadcom expects third-quarter AI semiconductor revenue of about $16 billion, representing year-over-year growth of over 200% but still below the average analyst estimate of $17.2 billion. More disappointing for the market was that Broadcom maintained its 2027 fiscal year AI chip revenue guidance of "over $100 billion" unchanged, failing to raise it significantly as the market had hoped. As a result, Broadcom's stock price continued to fall heavily.

Broadcom's weak guidance raised doubts about whether the AI investment boom is cooling. However, analysts generally believe the main reason for this round of selling is not a deterioration in fundamentals but rather that market expectations had become overly saturated.

The tenth-ranked stock was Intel (INTC), which closed down 11.28%, with a turnover of $14.497 billion. Hitachi and Intel announced a strategic partnership on June 5 aimed at accelerating AI transformation in key industries. The two companies will combine their strengths in IT, operational technology, and advanced computing, focusing on five pillars: foundry tools, quantum computing, custom chips and edge AI, and factory automation to modernize operations and upgrade infrastructure. This collaboration will prioritize improving robot yield rates and energy efficiency in manufacturing facilities, including Intel's own factories. Intel's fabs will deploy Hitachi's AI service "HMAX Energy" for core power equipment management, while Intel plans to supply high-voltage chips to Hitachi to optimize its power systems. Together, they aim to optimize semiconductor manufacturing processes using cutting-edge technologies like physics-informed AI.

The fifteenth-ranked stock was Taiwan Semiconductor Manufacturing (TSM), which closed down 6.69%, with a turnover of $8.035 billion. TSMC Chairman C.C. Wei stated at the shareholders' meeting on June 4 that in its 40 years, TSMC has never lacked competitors, and its approach to challenges has been to "work hard to keep beating them."

Addressing the potential threat from Intel's advanced packaging technology, Wei confidently responded, "I'm not afraid." Regarding Samsung's claim to surpass TSMC within 10 years, he bluntly remarked that Samsung has three times over the past 20 years claimed it would "catch up to TSMC in 10 years," but has always failed to deliver. "My evaluation for them is—dreaming."

The nineteenth-ranked stock was Qualcomm (QCOM), which closed down 10.98%, with a turnover of $5.257 billion. At the 2026 Qualcomm Automotive Technology and Collaboration Summit, Qualcomm Technologies, along with ecosystem partners including ArcSoft, China Unicom Smart Connection, Zebra Intelligent, Desay SV Automotive, Meijia Technology, and Thunder Software, announced the "Claw" in-vehicle AI ecosystem initiative.

According to the introduction, this initiative is based on a scalable modular architecture, combining the Snapdragon Digital Chassis solution with the Qualcomm AI Runtime environment, and leveraging partners' capabilities in cockpit systems, vehicle operating systems, AI agent middleware, AI applications, and mass production delivery. It aims to address the fragmentation issues hindering the development of next-generation automotive intelligence.

Through this initiative, Qualcomm Technologies and its ecosystem partners are committed to deploying AI agents and multimodal large models directly onto vehicles, driving the evolution of cars into intelligent companions.