semiconductors

A JPMorgan tech specialist, as reported by ZeroHedge, attributed today's tech selloff to a warning from Hock Tan about financing bottlenecks and renewed power concerns.

Microsoft plans multi-gigawatt deployments of its Maia 300 AI chips starting next year, betting custom silicon will defend cloud margins against rivals AWS and Google.

Intensified talks with state officials signal TSMC's commitment to spreading advanced manufacturing beyond its $65 billion Arizona complex and reducing geopolitical concentration risk.

Micron trades 18% below its June peak, but generative AI demand for HBM chips and enterprise SSDs is fundamentally altering the semiconductor's historical boom-bust pattern.

Supply constraints, not demand, are driving profit expansion across AI chips, memory and storage—and consensus price targets don't reflect the earnings upside ahead.

Market misread OpenAI's annualized revenue figures as demand weakness; on-chain and operational data suggest AI infrastructure spending remains elevated.

Multi-year deal secures advanced packaging components for AI chip production as foundries race to scale CoWoS technology.

Taiwan's chip foundry reported NT$511.86 billion in September revenue, up 54.6 percent year-over-year, signaling sustained AI and high-performance computing demand despite a slight sequential dip from August.

Nvidia projects 70 percent revenue growth next fiscal year, implying a forward valuation of 15 times earnings versus 30 times trailing—a gap suggesting 100 percent upside if the multiple re-rates.

The Shanghai chipmaker taps markets for capital as its stock swings 50 percent from June highs, raising questions about valuation discipline in a crowded AI hardware race.

Q3 delivery beat and $30B credit facility back in-house semiconductor push for autonomous vehicles and robotics

Broadcom arranges over $50 billion for OpenAI's custom silicon while Oracle seeks separate funding for major chip purchases to control AI infrastructure costs

CEO Lisa Su confirmed substantial 2027 capacity increases and tens of billions in global investments to meet sustained AI hardware demand, extending supplier planning windows to three to five years.

Marvell Technology raised its fiscal 2028 revenue forecast to $20 billion on AI chip demand; Seaport initiated a Buy rating with a $270 price target, citing the company's unique full-stack competitive moat.

Nvidia hits record highs on AI demand and Fed pivot. At 15x 2028 earnings, the stock trades cheaper than peers despite superior execution.

Scott Melker highlighted Nvidia's record close at $239.24, placing its market capitalization near $6 trillion, in a post on X.

CEO Lisa Su is locking down advanced wafer capacity and memory supplies to meet surging demand, but TSMC's limited output remains the bottleneck

TSMC discussions for Elon Musk's $16.8 billion Texas chip facility displace Intel as primary supplier candidate, triggering 4 percent Intel stock decline.

Micron and Samsung executives say HBM demand for AI will consume growing wafer capacity through the decade, starving consumer devices of affordable memory.

The S&P 500 fell 0.3 percent Wednesday as Treasury yields stayed elevated, but Nvidia, AMD, and Micron drove the Nasdaq 100 to a 3.2 percent monthly gain

The partnership releases TileLang, a new programming language for Huawei's Ascend chips, aiming to reduce developer friction and give Chinese companies an alternative to Nvidia's ecosystem.

AMD acquires Fei-Fei Li's startup to build world models into chip design, betting physically-grounded AI applications will drive higher-margin GPU demand than Nvidia's software push.

The startup's VCSEL technology targets a real bottleneck—data movement energy costs in GPU clusters—but faces entrenched competitors and the hard problem of customer qualification.

Data center strength and the Apple foundry deal mask an 81x forward multiple that makes a year-end $150 price target a long shot.

Revenue hit $41.46 billion in Q3 on HBM sales, but $7.1 billion quarterly CapEx and sold-out inventory through 2027 hinge on AI architecture stability.

Samsung's board is set to approve a massive shareholder return program today, with shares already climbing 13% over two sessions as investors position ahead of official details

Taiwan posted 12.92 percent real GDP growth in Q2 2026, driven by semiconductor sales, but debate intensifies over wealth distribution in chip-dependent economies

Semiconductor, biotech, and fintech stocks rally as Treasury yields fall and AI capital expenditure forecasts stabilize, with AMD, Moderna, and Robinhood entering technical buy zones.

The substrate maker plans NT$53.7 billion in capex to unlock high-bandwidth packaging as industry faces potential 42 percent supply shortfall by 2028

Micron's plan to triple high-bandwidth memory output by late 2026 should sustain gross margin expansion above 70 percent and deliver outsized earnings beats through fiscal year-end.

The Dutch startup is building processor and memory architecture designed for AI inference, targeting 2028 deployment with a dual revenue model: hardware sales and IP licensing.

The ex-a16z partner is betting that traditional venture capital missed the AI infrastructure play — and that compute access, not just money, is the moat.

Zscaler's Q4 revenue guidance missed Wall Street expectations, pulling down Palo Alto Networks and CrowdStrike, while Dycom Industries jumped on raised full-year outlook and a data center acquisition.

South Korean chipmakers reject Korea Electric Power's upfront payment proposal, signaling caution over semiconductor demand durability.

Nvidia's fiscal 2028 guidance of 70 percent revenue growth masks even stronger underlying demand, with analyst price targets ranging from $350 to $580 by January 2028.

Intel shares jumped 9 percent Tuesday on CPU price increase reports, High NA EUV manufacturing progress, and Trump administration signaling on government stake sales

Bank of America raises its price target 45 percent on strong high-bandwidth memory supply constraints and expanding long-term contracts

CEO Matt Murphy credits a decade of supplier trust with major tech companies and diversified AI partnerships as the engine behind Marvell's outperformance of Broadcom

Arm's new CSS for Mobile 2 platform pairs neural accelerators with GPUs to handle persistent AI agents within smartphone power budgets, signaling a shift toward on-device AI workloads over cloud inference.

New semi-custom platform on TSMC N3P doubles socket performance versus N3, reshaping the CPU economics for hyperscalers building custom silicon

CEO Lip-Bu Tan is raising PC CPU prices 10 percent in October and reviewing product lines for profitability, signaling a fundamental shift away from volume-driven strategy.

Japanese chipmakers and AI stocks led the rally, but yen volatility and bond yield concerns limited gains on the Topix Index

With 17 percent in Nvidia and over 60 percent in semiconductors, Vanguard's tech ETF has beaten the S&P 500 for four straight years on sustained hyperscaler AI capex.

Nvidia reported 106 percent sales growth to $96.2 billion in Q2 2027 yet trades at a 24.2 forward P/E—a disconnect that underprices multiyear AI infrastructure spending.

Nvidia projects Q3 gross margin of 74%, below consensus of 74.9–76.5%, citing manufacturing and memory costs with a sharper dip to 71–72% expected in Q4.

NVIDIA's record Q2 results underscore a structural divergence: semiconductor makers are capturing significantly more value than the large technology firms deploying AI, with Micron and Marvell up 220 percent and 185 percent respectively in 2026.

AVGO fell 5 percent after reporting tripled AI chip revenue but guiding to slower growth ahead, signaling demand normalization.

Strong results followed by stock decline would confirm summer correction risk, strategist warns, as semiconductor concentration in S&P 500 reaches dangerous levels.

Jonathan Krinsky flags 15 days of 3%-plus swings in 30 sessions as a setup that preceded double-digit losses in five prior cycles.

Citi bumps Texas Instruments to top semiconductor pick, citing its commanding position in AI server power delivery—29.8% upside from current levels.