The one-page fundamentals report for every US stock

Alibaba is accelerating its data-center expansion in Europe and the Middle East as part of a plan to build a 20-gigawatt global infrastructure by 2032 to support its AI push. It will launch its first cloud regions in Turkey, Finland and the Netherlands within 12 months and broaden its footprint in Malaysia, Germany, the United Arab Emirates, France and Hong Kong, strengthening localized cloud and AI services. The move pits Alibaba against Western rivals like Amazon and Alphabet in regions outside Asia and aligns with China's aim to share its AI advances globally. Alibaba has also unveiled new AI hardware and plans to train foundation models with 5 to 10 trillion parameters; Citi estimates the 20GW network could generate over $160 billion in external cloud revenue. The company had earlier opened facilities in Brazil and France and outlined a three-year AI investment plan of $53 billion in 2025.

SanDisk (SNDK) has locked in buyers for about half of this year’s output and two-thirds of next year’s under long-term contracts that include volume commitments, minimum guarantees, and price floors/ceilings. Management says margins stay attractive even at floor prices, aiming to shield itself from a future memory downturn. The move follows an August investor day outlining a new business model with 10 agreements across eight customers. The contracts reduce exposure to the spot market and underpin targets for fiscal 2028–2030 of mid-to-high-teens revenue growth and around 80% adjusted gross margins, plus a $14B buyback funded by a windfall. Sandisk has benefited from AI data-center demand, with data-center sales now about a third of revenue; quarterly revenue rose 372% year over year to $8.97B, and net income swung to about $7B. Yet memory cycles are volatile, and pricing power could erode if supply expands.

Amazon.com blocked Meta's Muse AI shopping agent from executing purchases on its retail platform, removing direct access to Amazon's product catalog and checkout systems. Rival retailers remain open to similar third-party AI shoppers, creating a split in how large ecommerce platforms handle external agents. The move aligns with Amazon's narrative of tight vertical integration and AWS-driven AI infrastructure, signaling a priority on its own AI stack and commerce funnel over rival agents. It underscores a broader push to shield customer data and margins within Amazon's retail ecosystem, while raising questions about openness to third-party agents and potential pushback from partners like Meta, Google, or Walmart. The policy decision could influence competitive dynamics, investor sentiment, and future AI shopping tools across ecommerce.

Qualcomm unveiled two flagship Snapdragon chips—Snapdragon 8 Elite Extreme Gen 6 and Snapdragon 8 Elite Gen 6—at its Maui Snapdragon Summit, each designed to run agentic AI on-device to reduce cloud compute costs. The move expands Qualcomm's strategy beyond smartphones into edge devices and data-center components, with a goal of broader premium-device AI adoption. Management projects non-handset revenue approaching $40 billion by fiscal 2029, while handset and auto sales remain key drivers today. The company frames a multi-flagship approach as essential to staying ahead in AI hardware and expanding AI innovations into more devices.

AMD briefly breached $1 trillion in market cap, closing near $609.65 with shares up over 8% after hitting a record high of $615.99. It became the fourth U.S. chipmaker to top $1T, amid AI demand shifting from GPUs to include general-purpose CPUs for agentic workloads. Meta's Muse app is cited as a catalyst, with CPUs tightening supply across data centers; Intel also reported strong demand. AMD projects server CPU TAM growth >35% annually to over $120B by 2030, with total data-center CPU opportunity up to $220B. Investors question durability of the rally amid potential normalization of AI demand and possible limits on the CPU opportunity; ongoing shortages and data-center spend would bolster AMD's case.

AI-exposed software roles are paying up as frontier-model productivity shifts the labor mix. Advertised pay for AI-exposed jobs has risen roughly 46% since 2021, while entry-level postings fell from 29% to 10% and senior postings rose from 22% to 47%. The premium mainly reflects experience directing AI, not coding juniors. Employers are cutting junior training, creating a hollowed-out ladder where only senior engineers reap the benefits of the AI surge. The broader economy remains resilient: unemployment about 4.1%, job openings high, and displaced coders finding work outside software; the information sector grew to $1.79 trillion in Q1 2026. RAND warns a fiscal risk from reliance on wages, and Indeed's chief warns of a vicious cycle if big platforms freeze new-grad hiring while raising senior bands.

NVDA posted Q2 revenue of $96.22B, up about 106% YoY, with Data Center at $89.02B (Networking +138%); Q3 guide is about $108B ±2%. The CUDA software moat and full-stack AI platform are cited as differentiators vs AMD and Broadcom. Hyperscaler AI capex is forecast at roughly $800B in 2026 and $1.3T in 2027; NVDA says it can supply only about 70% of stated demand through fiscal 2028, highlighting a supply constraint. The piece notes massive free cash flow, buybacks, and a strong balance sheet, while puncturing risk from ballooning AI-related memory commitments and receivables if AI growth slows. Despite not making a top-10 buy list, the author remains bullish and plans to accumulate more NVDA.

JPMorgan Chase and the Qatar Investment Authority sign a Memorandum of Understanding to create a $20 billion framework across public and private markets. The package includes a $15 billion global public-equities mandate for JPMorgan Asset Management and a $5 billion private-markets initiative focused on providing senior financing to established U.S. middle-market companies in industrials, services, healthcare and technology. JPMorgan aims to deepen its relationship with a major sovereign investor while expanding its institutional asset-management and private-market capabilities, leveraging its global platform, research and client relationships. The arrangement could generate incremental fee-based income as assets are deployed and broaden cross-selling across investment banking, markets, financing and asset management. JPM sees private capital as a meaningful organic growth opportunity amid abundant liquidity and private-equity dry powder.

A coalition of major banks including NatWest, Bank of America, ING, Capital One, ASB Bank, and Commonwealth Bank of Australia warns that AI-powered shopping agents create new risks for consumers around scams and data privacy, outpacing current protections. They publish a report with principles for agentic commerce—where AI tools select and buy goods on a shopper’s behalf—and plan to present regulatory proposals. Concerns include agents collecting card details on third-party sites, favoring payment methods with weaker safeguards, and risks of fraud, impersonation, and social engineering. Liability across the value chain is unclear. Proposals call for notifying customers when an AI agent is involved, clearer explanations of decisions, stronger data security, and preserved consumer autonomy with interoperable platforms. Regulators are tightening AI oversight in finance, while AI adoption in shopping grows unevenly among retailers and shoppers.

Muse AI by Meta sparked a rally in META stock, up about 11% and adding roughly $192 billion to its market cap; Zuckerberg’s net worth rose by about $13 billion. Muse topped Apple's App Store rankings and posted 2.8 million US/Canada downloads in 12 days, outpacing ChatGPT's early pace. Meta AI chief Alexandr Wang and analysts say Muse could unlock a massive enterprise AI market, boosting usage and creating new monetization avenues—advertising, subscriptions, and transaction revenue share in a multi-trillion TAM—potentially transforming Meta’s ROI on AI investments and market dynamics in AI infrastructure.

Meta's Muse AI assistant has surged to No.1 on Apple's U.S. free app chart, boasting a 4.9-star rating from over 25,000 reviews. JPMorgan keeps an Overweight rating with a $820 price target; Bank of America also rates the stock Buy with an $810 target. Evercore's Mark Mahaney estimates that if 1% of Meta's 3.6 billion users subscribed at $20 per month, annual revenue could reach about $8 billion, underscoring potential monetization beyond advertising. Analysts say Muse could become the most widely used consumer AI app since ChatGPT, leveraging Meta's distribution and product virality. Muse quickly topped the Productivity category post-launch and remains a signal of rising investor optimism about Meta's AI strategy. The discussion centers on multiple revenue paths—advertising, subscriptions, or revenue sharing—with user growth seen as the priority over immediate monetization. Meta's stock has rallied this year, amid AI buzz.

Apple unveils upgraded Mac Minis and Mac Studio desktops designed to run AI workloads locally, pitched as cheaper than renting cloud data centers. The machines aim at on-device AI tasks such as coding and business analytics, reducing the need for cloud tokens from OpenAI or Anthropic. Apple even demonstrated four Mac Studios linked together to run a trillion-parameter model off a single wall outlet, illustrating on-device scalability. Apple argues there’s no per-token cost, just upfront hardware, positioning Macs against Nvidia-powered Windows PCs in the enterprise AI race. Apple’s enterprise share is small (about 4.6% vs. Windows' ~91%), but its emphasis on power efficiency and tight hardware-software integration is seen as a potential shift. Microsoft stresses Windows ML and a future super app, while Nvidia remains strong in data centers. Real-world impact hinges on software, support, and cost economics at scale.

Broadcom could reach a $3 trillion market cap by end-2030, driven by management's forecast of accelerating AI semiconductor revenue. In fiscal Q3 2026, revenue rose 86% to $29.6 billion, with AI semiconductor sales at $16.7 billion (up 221% YoY); non-GAAP net income approached $16.4 billion. Management guided for about $34.8 billion in fiscal Q4 revenue, including $21.7 billion of AI semis. AI revenue is targeted at $58 billion for the current year, expanding to about $115 billion in 2027 and $230 billion in 2028, aided by secured supply. Combined with non-AI hardware, total revenue could approach roughly $280 billion in fiscal 2028. Margins are a key risk, potentially slipping if AI hardware mix widens, but the thesis hinges on delivering demand and supply as forecast.

Apple is reportedly developing enterprise AI servers built around future M8 Ultra processors, with designs using two or four chips and a possible 2029 launch aimed at AI inference. Nvidia NVLink Fusion could connect Apple’s processors, extending Nvidia’s interconnect technology to non-NVIDIA CPUs and accelerators. The plan isn’t finalized and may never ship, but it signals a shift in AI hardware economics: Nvidia could monetize networking gear used in Apple’s AI systems even as Apple owns the compute, while Apple gains a local-inference channel for developers, governments and businesses. Networking equipment might account for 10%–15% of AI data-center costs, widening Nvidia’s revenue beyond GPUs. Apple would be re-entering servers after Xserve’s 2011 exit, and a 2029 target provides room for competitors to respond. Hedge funds show interest in both AAPL and NVDA, though positions remain mixed.

Arm Holdings shares jumped 17% after CEO Rene Haas said demand for Arm-based technology is 'off the charts' and signaled growing confidence Arm can exceed a $2 billion revenue target for its developing AGI CPU business, up from an initial $1 billion target and then $2 billion later in the year. The pipeline was reportedly above that level by September. Arm’s chips underpin competing AI CPUs from Nvidia, Google, Amazon and Microsoft, aligning with a broader shift to CPUs bundled with AI accelerators. Haas also warned of long lead times due to wafer, substrate and testing capacity constraints, underscoring strong demand and bottlenecks. The company is seeking more value from chips beyond licensing, and the key test is turning AI demand into revenue and profit to justify higher expectations.

Shopify enabled Meta Muse across all stores via Shop Pay, while Amazon blocked Muse, creating a bifurcated e-commerce landscape. Shopify’s move follows its 2025 launch of Agentic Storefronts, making US stores discoverable by AI agents by default and auto-sharing catalogs. In Q1 2026, AI-driven traffic and orders on Shopify surged, signaling growing agent-driven commerce. But PYMNTS testing in September showed friction: Muse’s checkout used Stripe Link and required multiple credential approvals before tasks could be completed, far slower than manual shopping. The split deepens: Shopify’s integration promises seamless discovery and checkout; Amazon’s gatekeeping preserves marketplace control. Independents rely on Stripe Link or browser automation, lacking native, end-to-end flow. Broader pushes by Visa, Mastercard, and Ant Group toward agent identity frameworks point to an industry-wide shift toward agent-based shopping.

Silver futures opened at $66.51/oz on Sept 22, 2026, up 0.2% from Monday, after peaking at $67.32—the highest since Sept 9. Prices are driven by industrial demand for AI hardware and solar panels. Key near-term driver is the Trump-Xi meeting this week, with tariffs and AI controls expected to be on the table. Chinese export curbs on critical minerals and rare earths loom as potential supply shocks for US and global producers. Analysts forecast mixed paths for silver, with some predicting prices above $80 by year-end and possibly $100 by 2030, while near-term volatility remains possible. The latest price chart shows some declines from intraday highs, reflecting ongoing volatility and shifting sentiment around the talks and macro demand.

Visa has moved from experimenting with stablecoins to operating blockchain infrastructure by joining Circle's Arc L1 as a validator. Alongside BlackRock, DTCC, Mastercard, and ICE, Visa is shifting from routing payments to securing the network itself. Stablecoin settlement volume grew from about $3.5B annualized in Nov 2025 to $7B in Apr 2026 and $20B by Sep 2026, signaling rapid production-scale utility. Visa uses a dual-chain approach: Solana handles production settlement in a weekly cycle with Cross River Bank and Lead Bank; Arc L1 is a permissioned, USDC-native environment for financial markets and AI-enabled economic activity. The move aims to provide compliant, high-trust rails ahead of GENIUS Act rules, with Arc designed for real-time settlement and liquidity. The shift marks a structural change toward regulated, enterprise-grade blockchain infrastructure.

Meta's Muse app surged to 730,000 downloads in five days, surpassing ChatGPT to become the top free iOS app in the US. By Sept 21 it had 2.5 million cumulative downloads (1.5M iOS, 1.1M Android). Muse's fast uptake outpaced Claude (400k) and Grok (200k). Meta stock jumped about 11.4% on Sept 21 to $741.25, aiding a Nasdaq close at a record 27,122.09. Bernstein says potential for broader adoption but warns on privacy and data-hungry integrations with Google Workspace, Spotify and Apple Health. Muse Connectors and a macOS app aim to extend usage; Amazon blocks Muse shopping; Shopify partners for agentic commerce. Muse pricing includes free tier, $20/month Muse Power, and $100/month tier; token costs $1.25–$4.25 per million. Sustainability hinges on translating velocity into ongoing usage.

As China and the US race to build separate AI supply chains, investors hedge by betting on both sides. Wall Street banks underwrite Chinese AI listings and follow-ons, while Chinese money flows into US tech, keeping deep financial ties even under chips and investment restrictions. Through mid-September, US-focused AI and chip deals accounted for about $17.2 billion of issuance, nearly 30% of sector activity, and US stocks remain top targets for Chinese outbound funds. US holdings worldwide total around $750 billion, up 23% in a year. Analysts warn a bifurcated AI ecosystem risks widening if tensions flare, with two internets and two chip stacks. Yet many investors see value in cross-border exposure as a safety net, and discussions on bilateral AI dialogue are under way. The outcome likely won’t produce a single winner.

Meta's Muse AI agent sparked a rally, with META stock up about 11% and billions added to market cap and Zuckerberg's net worth. Muse topped app-store charts with 2.8 million US/Canada downloads in 12 days, surpassing ChatGPT's early traction. Analysts say Muse could unlock a new, enterprise-focused revenue stream for Meta, leveraging advertising, subscriptions and transaction revenue across large TAMs. Evercore ISI's Mark Mahaney argues the market has over-focused on enterprise AI and Muse could prove a game changer if it penetrates business users. Meta has invested more than $200 billion annually in AI, and Muse offers tangible ROAI signals and potential monetization. If successful in the enterprise AI market, Muse could boost engagement and create a substantial profit driver beyond advertising.

AMD's market value topped $1 trillion as it cements its position as Nvidia's rival under CEO Lisa Su. Since 2014, Su has rebuilt AMD around data-center CPUs and AI accelerators, closing deals like Xilinx and striking major partnerships. In Q2, AMD beat estimates with non-GAAP EPS of $1.66 and revenue of $11.5 billion—up 50% year over year—with data-center sales more than doubling to $6.7 billion on EPYC CPUs and Instinct AI accelerators. The company projects 2H26 server CPU revenue up around 80% year over year and 2027 data-center revenue more than doubling, with AI GPUs rising well over 100%. New Instinct MI450 GPUs and 6th‑gen EPYC Venice CPUs underpin its Helios rack-scale AI system. Microsoft will roll Helios into Azure AI in H2 2026, and AMD plans an Anthropic investment—up to $5 billion and up to 2 GW of MI450 deployments—starting 2027, signaling a bold AI infrastructure push.

Nvidia’s stock trades at less than 17x forward earnings, the cheapest in more than a decade, despite multiyear growth expectations. The multiple has fallen from about 25x earnings as recently as May, with investors skeptical that Nvidia’s current earnings power is sustainable even as revenue and net income are forecast to surge (roughly 90% and 99% in fiscal 2027). Management highlighted 2028 sales growth around 70%, reinforcing a powerful growth narrative, but margins are expected to shrink to below 72% in Q4 before a rebound. Competition from memory makers and in-house chip initiatives by large customers could erode pricing power. The market remains wary that AI capex may slow or that regulatory changes could bite, even as AI demand and a broader semiconductor rally provide support.

Intel says it can meet only about 50% of current AI workload demand, signaling tight supply as Meta's Muse AI workloads drive stronger CPU orders. Rival CPUs are seen with rising AI-related demand, pressing capacity planning across upcoming production cycles. The spike highlights a broader AI-infrastructure supply squeeze where capacity constraints and older lines could limit actual shipments. Muse-driven demand reinforces the case for heavier investment in AI servers and foundry capacity, while investors will watch for data center volumes and any easing of CPU backlogs. The article notes this amid a broader 88-stock AI infrastructure landscape, with Intel leveraging its Xeon/Core Ultra footprint. A positive signal not captured here: stronger foundry traction and demand visibility could justify continued investment.

Armenia is positioning itself as a top hub for American AI infrastructure via Firebird’s Armenia data center in Hrazdan, near Yerevan. The facility aims for 300 MW capacity and more than 70,000 Nvidia chips by the end of next year, with about 20% of computing power reserved for domestic use and the rest for foreign firms including Perplexity AI. If completed, it would be among Europe’s largest computing clusters, fueling innovation in Armenia and nearby regions in the South Caucasus and Central Asia as powers vie for AI influence along a Europe-China corridor bypassing Russia. The project intersects geopolitics and export diplomacy: Trump-era chip-export licenses have been used to advance peace talks with Azerbaijan, while Armenia has tightened controls on electronics shipments to Russia. Nvidia highlights Armenia’s tech talent; expansions to Kazakhstan are also considered; a US-Armenia AI partnership underpins the effort.

Alibaba unveils the Zhenwu V900 AI accelerator from its T-Head unit, claiming triple performance and scalable clusters of up to 500,000 for frontier-model training. The company plans more than $53 billion in AI hardware and infrastructure over three years and recently raised about $10.2 billion in Hong Kong. CEO Eddie Wu targets 20 gigawatts of Alibaba Cloud data-center capacity by 2032 to meet rising AI demand, and the firm aims to develop a 5-10 trillion-parameter model. The push heightens competition with Nvidia in the AI accelerator market and signals a major expansion of Chinese AI compute ambitions, even as US labs warn about pace. Announcement comes ahead of a US-China summit amid renewed AI stock optimism.

Tech-led rally lifted Asia stocks while U.S. equities were mixed, as the Nasdaq closed at a record high on gains from Micron and other AI-linked names. The Nasdaq Composite rose about 0.45%; the S&P 500 was flat and the Dow slipped about 0.36%. Global shares inched higher as oil drifted lower on resumed Middle East flows; Brent and U.S. crude fell amid pipeline restart and Hormuz movement. Treasuries ticked lower in yield, but the 10-year yield remained elevated as traders priced in further central-bank tightening. The dollar edged up modestly against the euro and yen. Investors eyed a Trump-Xi meeting for signs of progress on a trade truce and possible AI collaboration, while warnings of further rate hikes kept policy risk in view. The BoJ's 31-year high rate decision also shaped expectations.

Palantir CEO Alex Karp argues AI leaders may never IPO, suggesting the industry should 'nationalize us, please' to curb liability from destructive cognitive‑computing tech. In CNBC remarks, he said liability belongs in government documents and that the first defense is firm liability, implying regulation is necessary if firms won't police themselves. He also urged the Fed to craft civil and criminal penalties for irresponsible ML development. The comments also spotlight Palantir’s interest in AI sovereignty and oversight, hinting at public‑sector solutions. The piece notes OpenAI’s past talks of Washington taking a stake and public wealth funds, as part of broader debates about regulation, nationalization versus IPOs, and how policy could reshape the AI arms race.

AI chip and memory shortage is driving divergent risk/reward between NVIDIA and Micron. NVIDIA backs buyers with balance-sheet financing and faces higher memory-price pressures, guiding a 70% revenue-growth outlook for fiscal 2028 and a steady margin around 72%–73% as prices rise. Micron secures buyers with about $22 billion in deposits and near-five-year take-or-pay contracts, with roughly $18B cash returned later; management says floors keep gross margins well above prior cycles, even with a price cap near mid-2026. Micron has 16 strategic agreements, 14 of which carry minimum-revenue floors around $100B, potentially delivering more stable near-term cash flow. On valuation, Micron trades cheaper on EBIT and shows stronger recent growth, while NVIDIA has the longer operating history and bigger scale. The verdict favors Micron’s safety vs NVIDIA’s financing risk, even as both benefit from AI demand.

Eli Lilly's oral GLP-1 pill Foundayo (forglipron) hit U.S. pharmacies in April and posted $98 million in sales in its first full quarter, while Novo Nordisk's Wegovy logged about $494–$497 million in the same period. Lilly says Foundayo now accounts for about one-third of new GLP-1 tablet starts and is expanding as it competes for market share. Lilly broke ground on a $6.5 billion Houston plant to produce Foundayo and other small-molecule drugs, part of a broader plan to onshore manufacturing with four plants and a $27 billion pledge by 2025. Foundayo is a once-daily, non-peptide GLP-1 receptor agonist that can be taken without fasting, broadening Lilly's obesity-drug portfolio beyond injections. Medicare obesity-drug coverage starting July could widen access to Foundayo and Zepbound. LLY stock rose about 1% after the news.

AMD briefly topped $1 trillion in market value for the first time, peaking at $615.99 per share before retreating to around $609.65 for about $995 billion. Nvidia rose 2.3% that session; Intel jumped 12% to $121.67 and Arm 15%. AMD thus became the fourth U.S. chipmaker above $1 trillion, joining Nvidia, Broadcom and Micron. The piece cites three rally drivers: heavy CPU demand to run AI inference alongside GPUs; AMD warning of up to ~10% price hikes on AI accelerators, GPUs and chipsets, boosting margins; Nvidia’s scale complicating the rally; and Q4 results to show whether price gains translate into income or merely lifted demand.

St. Louis Fed President Alberto Musalem says the Fed will likely need to hike rates further to curb inflation, which remains above target due to strong demand and supply shocks. He argues earlier, incremental tightening is better than delayed, larger action. Inflation persists even after stripping oil and other factors; PCE inflation was 3.7% in July, well above the 2% target. Commodity prices—diesel, copper—have climbed amid broader supply concerns and an AI-investment boom, while US domestic spending stays resilient. The Fed's policy rate at 3.75%-4.00% is still accommodative, and markets expect roughly three more quarter-point hikes through April. Musalem cautions the labor market is not the inflation source, and he did not specify a target rate level. He emphasizes needing meaningful restraint to reach the inflation target in about 18 months.

Kairos Power secures Samsung C&T as an equity investor and in-kind engineering partner in a deal worth up to $100 million to build Kairos’ 50 MW demonstration reactor for Google, aiming for completion around 2030. Samsung will invest about $70 million and supply $30 million in engineering services. The arrangement supports Kairos’ Hermes program: Hermes 1, a low-power demonstrator under construction in Oak Ridge, and Hermes 2, the first commercial-scale unit whose output will count toward Google's roughly 0.5 GW deal. The project uses a fluoride salt–cooled high-temperature reactor design with TRISO fuel, an approach intended to reduce meltdown risk. NRC approval was earned in 2024, and the timeline could accelerate subsequent reactors as data-center demand rises.

Arm is shifting from royalty-based income to in-house silicon manufacturing, aiming to reshape chip economics. Improved supply confidence helped justify a surge in demand signals—up to $2 billion in AGI CPU demand—while margins on in-house silicon are much lower than Arm's 92.5% royalty margin, making the mix shift meaningful. The company faces capacity as the decisive variable; first shipments at the end of calendar 2026 are viewed as essential to sustain valuation. Nvidia's Vera CPU feeds Arm royalties, and Qualcomm's Nuvia license trial in Q4 2026 could reprice per-chip royalties across the industry. Arm's stock has danced on demand figures, and a 24/7 Wall St ranking excluded ARM from its Top 10 Stocks to Buy. The bull case hinges on AGI CPU TAM potentially nearing $200 billion, with Neoverse shipments exceeding 1.5 billion cores.

AMD briefly surpassed $1 trillion in market capitalization, joining Nvidia, Broadcom, and Micron as the only U.S. chipmakers to reach the milestone. The gain followed investor bets on AI computing and AMD’s expansion from chips to integrated systems that combine processors, networking gear, and related hardware. Shares rose to about $610, an all-time high of $613.92, as the company tightens competition with Nvidia in GPUs and gains server market share aided by demand for CPUs used alongside graphics processors for inference. AMD previously forecast revenue above estimates but guided below investors’ expectations, triggering a stock pullback before a subsequent rally that lifted the stock more than 26% since the update. Nvidia remains far larger, with a market cap above $5 trillion.

Joby Aviation completed its first fully autonomous cross-country flight across the United States, flying 3,199 miles with zero control inputs from a safety pilot. The converted Cessna Caravan executed taxiing, takeoff, navigation and landing under remote supervision from Joby’s Autonomy Headquarters in California and Shaw Air Force Base in South Carolina, up to 2,323 miles away. The voyage, part of the 2026 Electric Skies tour, tested autonomy in diverse environments and weather, including Phoenix Deer Valley airport, with real-time rerouting. Stops included Concord, CA; Phoenix, AZ; Fort Worth, TX; Shaw AFB; and Outer Banks, NC, with a return leg planned through Raleigh, Fredericksburg/Washington, Louisville, Wichita, Oklahoma City, Salt Lake City and Portland. Built on the Xwing autonomy stack, the system has more than 400 flights and 800 automated hours. Joby aims to scale autonomous air logistics by merging autonomous systems with certified airframes.

Advanced Micro Devices crossed a $1 trillion market capitalization for the first time as its stock surged on a five-session rally fueled by AI chip demand. The intraday high reached $613.92, placing AMD among Nvidia, Taiwan Semiconductor, Broadcom and Micron in the exclusive trillion-dollar club; 15 other public companies are at that level. The five-session run adds about 24% and lifts the year-to-date gain beyond 180%. Nvidia remains the AI-chip leader with a roughly $5.4 trillion market cap. The rally broadened to Intel and Arm as the Philadelphia Semiconductor Index climbed. Investor enthusiasm ties to the rise of Meta’s Muse AI Agent, which boosted demand for processors to run agentic workloads; Meta is AMD’s second-biggest customer at about 5.5% of revenue. AMD’s AI-focused data-center business led growth: Q1 revenue $10.25B, data-center $5.78B; Q2 revenue $11.54B, data-center $6.7B. AMD targets doubling data-center revenue by 2027.

Iris Energy holds $7.6 billion in cash after a major equity raise to fund a 510‑MW Childress AI data‑center in Texas, part of a pivot from Bitcoin mining to AI infrastructure. The project targets hyperscale customers with long-term power contracts as it continues mining operations in parallel. Q4 2026 revenue rose to $71.5 million and hash rate reached 31 EH/s, but FY2026 operating income swung to a large loss due to noncash depreciation and impairment tied to the buildout. The cash hoard lets the company complete Childress without dilutive financing, though near‑term results remain sensitive to construction timing, power availability, and customer deployments. Long‑term valuation scenarios imply meaningful upside if ramp goes as planned, but execution risk and delays could weigh on near-term returns.

Chip stocks led markets as the PHLX Semiconductor Index jumped more than 3%. AMD surged to an all-time high above $610, briefly pushing its market cap to $1 trillion. Arm Holdings and Intel bounced about 10% as Meta's AI agent MUSE spurred a rally among server CPUs. Nvidia edged higher as demand for AI hardware remains robust. Intel has rallied more than 35% over the past month; AMD and Arm up roughly 30%. Analysts note tailwinds from ongoing data-center buildouts, AI inference demand, and talk that SK Hynix is exploring a deal with Intel. Bank of America's Vivek Arya lifts his 2030 chip-market estimate to $3.2 trillion, citing hardware and power constraints and rising demand for memory and logic. Still, markets expect little sign of a cap on AI spend despite earlier concerns.

Alphabet’s Google launched Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two voice-AI models designed to act as capable enterprise agents. 3.8 Live can execute tools and API calls within dialogue, process visual information in near real time, and operate in 97 languages; Extended Thinking handles multi-step reasoning for complex tasks. Google is collaborating with Salesforce, Genspark, and Lumeris to embed the tech into enterprise workflows. The Gemini app has about 950 million monthly active users, and Gemini models process 22 billion API tokens per minute. Management raised 2026 capex guidance to $195-205 billion to rapidly expand capacity. Google Cloud posted strong growth (revenue up 82% YoY to $24.8B; operating income $8.8B). Analysts are optimistic, with price targets around $432, suggesting meaningful upside if AI investments translate into monetization.

Novo Nordisk set long-term growth targets ahead of semaglutide patent expirations; US-listed shares declined after the plan was unveiled, signaling investor concern about growth as key products lose protection.

Meta Platforms posted Q2 2026 revenue of $60.8B, up 28% YoY, with ad impressions up 14% and price per ad up 12%; daily active users reached 3.6B across Facebook, Instagram, WhatsApp, Messenger and Threads. Zuckerberg said the ads business is growing faster on a dollar basis than any other large ad platform. For 2026, capex guidance rose to $130–$145B to power AI and infrastructure, nearly double 2025’s spend; Family of Apps generated about $115.8B in operating cash flow in 2025. Q2 free cash flow was $784M. Reality Labs losses totaled $19.2B in 2025, with a $4.03B Q1 loss in 2026 and $2.4B in Q2 charges. Alphabet posted negative free cash flow and suspended buybacks; Snap trades down about 93% in five years. The bet remains that Meta’s ad engine will keep growing despite high capex, with a ~24x earnings multiple. Meta wasn’t in 24/7 Wall St’s top 10.

London-based AI hyperscaler Nscale filed for an IPO, the first major AI listing since safety debates broadened, with Anthropic cited as a key customer. The S-1 frames AI financing as circular and highlights Nvidia. Renaissance Capital estimates around $2 billion in potential proceeds. Spun out of a crypto-miner two years ago, Nscale posted a $1.02 billion net loss on $140.6 million in revenue for H1 2026, up from a $369 million loss on $10.4 million revenue a year earlier. The company reports over $103 billion in total contracted value and has raised billions in VC funding and debt, most recently at a $13.5 billion post-money valuation from investors including Aker, Sandton Capital Partners, and Paladin Capital. The filing underscores rapid AI growth and evolving financing dynamics in the sector.

Paramount Skydance and Warner Bros Discovery have reached a settlement with California and other states over their $110 billion merger, potentially clearing it to close one of history’s largest media deals. The agreement includes concessions such as a roughly $1.5 billion California production investment, commitments not to sell studio lots, penalties if Paramount fails to produce about 30 films a year, and measures to preserve editorial independence for CNN and CBS via independent boards. Talks also contemplated selling some cable channels. A $7 million daily fee for delays past September 30 would be discharged. EU and UK approvals are already in place, and Bloomberg described the settlement as removing a key obstacle as investors pushed Warner Bros Discovery and Paramount Skydance shares higher.

Amazon has blocked Meta's Muse AI assistant from making purchases on its site, escalating a dispute over who controls the online shopping experience when AI acts on a customer's behalf. A popup now appears warning that continued access by an unauthorized AI agent violates Amazon's Conditions of Use. Amazon said it had approached Meta and asked that Muse be kept out of its shopping flow, arguing third-party apps that buy from other businesses should operate with the merchant's consent and that Muse could conceal its identity and access account details without Amazon's knowledge. Meta launched Muse to perform emails, travel bookings, form filling, and purchases, using a dedicated cloud VM and Stripe Link to generate one-time card numbers; Sentinel monitors allowed actions. Muse is available on iOS, Android, web, and WhatsApp, with free and paid tiers. Meta says Muse has no visibility into passwords or payment methods; credentials go to secure storage.

Bitcoin traded above $85,000, nearing an eight-month high after recovering from a year-long drop of up to 34%. A friendlier regulatory backdrop and renewed investor interest fuel the rally, though near-term calls remain difficult. Analysts see support above $75,000 potentially opening a path to $82,000 and then to $90,000, with Bitcoin’s weekly chart and moving averages adding to bullish momentum. Yet critics warn cryptocurrencies do not resemble traditional earnings, and the broader global impact is uncertain. The piece notes Trump-era policy shifts improving the environment for crypto and the GENIUS Act creating a framework for stablecoins. It also cites caution from Motley Fool on Bitcoin as an investment, while acknowledging the longer-term appeal of crypto and the ongoing innovation in networks like Ethereum and Solana.

Morgan Stanley warned the S&P 500 could fall to about 7,100 if tighter financial conditions and higher energy prices push yields toward 5%, making equities less attractive. Diesel prices are driving persistent inflation as roughly 70% of U.S. freight moves by truck, spreading price pressures through manufacturing and consumer costs. Oil trajectory remains a risk to inflation and the broader economy, especially with potential Middle East disruptions. AI stocks face two-sided risks: a major security breach or reduced corporate AI spending could erode momentum, despite prior enthusiasm. The piece promotes a free report listing the ‘Top 10 Stocks To Buy Now’ from 24/7 Wall St.

Micron Technology has surged about 256% in 2026 and is set to report fiscal Q4 2026 results on Sept. 30. AI infrastructure demand supports memory-chip sales, as DRAM remains in a supply shortfall. PwC projects annual AI data-center capex at $800 billion in 2026, rising to $1.8 trillion by 2050, with AI hardware spending totaling $31.6 trillion by 2050. Goldman Sachs notes DRAM demand could outpace supply by about 5% in 2026 and 5.9% in 2027, with volume from new factories delayed until 2029. Consensus estimates put Q4 revenue around $50-51 billion and about $31 per share, with some forecasts calling for $56.9 billion and $34.95. If earnings beat and guidance improves, Micron could get a valuation boost and an outsized stock move post-earnings, aided by continued AI infrastructure growth and a tight memory market.

Novo Nordisk fell about 7% after presenting a long-term growth plan aiming for more than five multi-blockbuster launches by 2030 and growth in line with peers. Investors focused on a slower-growth posture and the looming semaglutide patent cliff in 2032, threatening a large share of sales. The move appears company-specific rather than a sector-wide shift, as Lilly slipped slightly, Viking Therapeutics rose on next-gen obesity hopes, and SPY rose while XLV dipped little. Novo Nordisk pledges diversification beyond obesity and diabetes into liver, cardiovascular and other diseases. Market reaction implies valuation re-pricing for future growth rather than current earnings as investors digest the day and assess where shares floor and ceiling may lie.

Anthropic reportedly eyeing a multitrillion-dollar IPO as SpaceX taught investors valuable lessons. SpaceX priced at $135 with an IPO pop to as high as $176–$225, but long-run returns have not rewarded early buyers; many are underwater. The piece outlines three takeaways for mega-IPOs: IPO price reflects demand, not intrinsic value; lockups and insider selling can press early prices; and waiting for post-IPO volatility can lead to a better entry. For Anthropic, the article notes a $2 trillion+ valuation target, a claimed $65 billion annual revenue run rate, plus large commitments to hyperscalers and chipmakers, and potential customer concentration with hyperscalers. The piece suggests examining the S-1 for revenue quality, commitments, and concentration, and cautions against buying on the first day, recommending waiting for the dust to settle to form a valuation based on fundamentals.

Shares of Greenland Energy (GLND), Greenland Mines (GRML) and Critical Metals (CRML) surged after a new U.S.-Greenland-Denmark security framework signaled a larger Washington role in Greenland’s defense and security, prompting expectations of greater scrutiny and potential funding for Arctic resource projects. Greenland Mines is advancing rare-earth and precious-metals developments; Critical Metals is developing the Tanbreez deposit; Greenland Energy is pursuing oil and gas in eastern Greenland. No new funding, permits, or commercial agreements were disclosed. The pact, described as preserving Greenland’s sovereignty, is expected to finalize during the United Nations General Assembly. Investors linked the framework to increased strategic importance of rare earths and other critical materials.

Futures rise as AI stocks rally and crude slides. Nasdaq closes at a record led by AMD, with Intel up 12.2% and Arm 17%; PHLX semiconductors up 4%. Nasdaq and S&P 500 gain, with the S&P around 7,765 and the Dow near 52,049. Meta climbs after a price-target upgrade linked to Muse AI. The 10-year yield dips below 5% and Brent falls under $100. Traders weigh potential UN diplomacy breakthroughs and talk of extending the U.S.-China trade truce. Bitcoin jumps over 6% to a seven-month high. Greenland-related stocks rally on negotiations. Market volume sits around 16.5 billion shares as AI-driven optimism meets evolving rate expectations and geopolitical headlines.

Broadcom CEO Hock Tan pushed back on AI-slowdown fears, insisting long-term AI revenue targets remain intact after Anthropic emerged as its likely largest custom silicon customer. Broadcom guided AI semiconductor revenue to $115 billion in fiscal 2027 and $230 billion in fiscal 2028, with Anthropic contracted to deploy 5 gigawatts of its TPU v8i chips in 2027 and potentially 10 more in 2028. The market responded with Broadcom stock dropping about 4.8% on Sept. 14 as semis slid, despite Tan's comments. Past calls have sparked volatility: in June, Broadcom fell after guiding no upward revision to a $100 billion AI target. Anthropic’s commitments provide visibility that may insulate Broadcom from pure pacing concerns, but investors remain sensitive to guidance gaps and Anthropic’s capex plans. Insider Monkey cautions volatility and focuses on Anthropic orders as a real signal.

Buffett is stepping down as Berkshire Hathaway chairman; Howard Buffett becomes non-executive chairman while Greg Abel takes over as CEO and head of capital allocation. Berkshire's blend of insurance float, long-term ownership, and a concentrated public-equity portfolio (Alphabet, American Express, Apple, Bank of America, Coca-Cola) helped deliver exceptional long-run returns: 1965–2025 per-share value rose about 19.7% annually, with a massive overall gain. The balance sheet is gigantic: roughly $1.26 trillion in assets, $750.2 billion in equity, $323.8 billion in equity securities, $359.2 billion in cash and Treasuries, and around $176 billion in insurance float. In 2025, operating earnings were $44.5 billion; 2026 first-half results included $35.8 billion net profits and about $11.4 billion after‑tax investment gains. Berkshire has deployed cash into deals like OxyChem ($9.4B) and Taylor Morrison ($6.8B). Past misses (Dexter Shoe, Kraft Heinz) underscore the importance of price and quality; Abel’s challenge is how to deploy vast cash going forward while preserving culture.

Retail investors are rushing to gain exposure to Anthropic and OpenAI before their IPOs through SPVs and high-fee funds. Brian Smith, 66, spread across eight brokerages and is seeking hundreds of thousands of dollars in shares, potentially 30% of his net worth. Exotic funds like Fundrise’s VCX, ARK’s ARKVX, and Robinhood Ventures Fund I/II promise private holdings or contractual exposure, but come with lockups, opaque ownership, and hefty fees. SpaceX private-market lessons linger: liquidity can evaporate and apparent gains can reverse. Some investors fear hype-driven bets; others see a rare opportunity given private-market access and big AI valuations. Anthropic reportedly targets a $2 trillion IPO; OpenAI is weighing a $1+ trillion round. Private exposure is accelerating as companies stay private longer, amplifying risk and potential volatility for retail investors and markets.

Nvidia has dominated the AI rally, but the next phase may favor software-and-security firms instead of chips. The iShares Expanded Tech-Software Sector ETF (IGV) is highlighted as a potential beneficiary if AI development slows, boosting demand for software and cybersecurity products. The piece notes a shift in leadership from megacap semis and memory to software as investors reward concrete results over hype. In 2026's first half, semiconductors surged while software lagged; in the second half, IGV rose about 15% as SMH slipped. Cybersecurity names such as CrowdStrike, Cloudflare and Zscaler are cited as beneficiaries of higher security spending. The Motley Fool also flags software strengths, arguing Act 2 of AI could be bigger than Act 1, potentially reshaping market dynamics and investor sentiment.