Highlighted Companies
KEY COMPANIESNVIDIA reported a record quarter with revenue of $81.62B vs. $78.86B expected, driven by data center revenue of $75.2B. Adjusted EPS of $1.87 topped the $1.76 consensus. The company announced an $80B share buyback program and raised its quarterly dividend 25-fold. Despite the blowout results, shares dipped post-earnings — the fourth consecutive post-earnings slide — as investors weighed whether the AI infrastructure cycle has been fully priced in. CEO Jensen Huang described a “five-layer cake” AI economy spanning energy, chips, infrastructure, models, and applications.
Read More →CEO Jensen Huang told CNBC that NVIDIA has “largely conceded” the Chinese AI chip market to Huawei due to U.S. export restrictions, and has told investors to “expect nothing” regarding approvals to sell advanced chips into China. China once accounted for 20%+ of NVIDIA’s data center revenue. Huawei had a record year and is expected to have “an extraordinary year coming up.” The comments underscore how Washington’s restrictions have accelerated Beijing’s semiconductor self-sufficiency push, permanently reshaping the competitive landscape.
Read More →AMD announced a $10B+ investment across Taiwan’s semiconductor and AI ecosystem, focusing on advanced packaging and manufacturing partnerships with ASE, SPIL, Sanmina, Wiwynn, Wistron, and Inventec. The investment supports Helios, AMD’s next-generation AI server system deploying in H2 2026. CEO Lisa Su said “agents are driving tremendous demand in the AI cycle.” AMD shares have doubled YTD as the company positions itself as the primary alternative to NVIDIA in the accelerator market, with 2nm EPYC processors now entering production at TSMC.
Read More →NVIDIA shipped its first Vera CPU systems to leading AI companies including SpaceX, Anthropic, and OpenAI as part of its broader push into next-generation infrastructure for agentic AI workloads. The Vera CPU platform complements NVIDIA’s Blackwell and Rubin GPU lines by targeting the growing CPU-intensive portion of AI inference — particularly agent orchestration, planning, and tool-use layers that don’t require raw GPU throughput. The move positions NVIDIA to capture more of the full-stack AI compute dollar.
Read More →AI & Semiconductors
AI & CHIPSGoogle launched Gemini 3.5 Flash at I/O 2026 as the default model across the Gemini app, AI Mode in Search, and developer tools. The model surpasses 3.1 Pro in coding, agentic, and multimodal benchmarks while running at 4x faster output tokens per second — effectively delivering frontier intelligence at Flash-tier cost and speed. Google also announced Gemini Spark, a general-purpose AI agent that can reason across connected apps and take action on users’ behalf. Gemini 3.5 Pro arrives in June. The launch intensifies the three-way model race between Google, Anthropic (Claude 4), and OpenAI.
Read More →Anthropic is closing a ~$30B fundraising round at a $900B pre-money valuation — surpassing OpenAI’s $852B round from March to become the world’s most valuable private AI company. The company projects Q2 revenue of $10.9B (doubling from Q1) and expects to achieve its first profitable quarter. Annualized revenue is on track to exceed $50B by end of June, up from $4B just 11 months ago. The velocity of Anthropic’s revenue growth — driven by enterprise Claude deployments — validates the thesis that enterprise AI, not consumer, is where the monetization war is being won.
Read More →OpenAI has filed confidential IPO paperwork with Goldman Sachs and Morgan Stanley, targeting a listing as early as September 2026 at a valuation above $1 trillion. The company’s last private round valued it at $852B. The IPO arrives amid questions about OpenAI’s path to profitability — the April miss on revenue targets still looms — but the filing signals confidence that public market appetite for frontier AI exposure will overcome unit economics scrutiny. If both Anthropic ($900B private) and OpenAI ($1T+ public) achieve their targets, the frontier AI lab sector alone will exceed $2T in combined valuation by year-end.
Read More →Markets & Tech Stocks
S&P 500 · NASDAQ · MARKETSThe S&P 500 rose 0.37% to 7,473.47 and the Dow gained 294 points to a new record close of 50,579.70 on Friday, capping an eighth consecutive winning week for the S&P — the longest streak since December 2023. The Nasdaq added 0.19% to 26,343.97. Treasury yields eased with the 10-year settling around 4.56% after hitting a one-year high mid-week. Geopolitical optimism around Iran peace negotiations provided a tailwind, while oil prices moderated. The rally breadth expanded with health care posting its best week since November alongside semiconductors.
Read More →NVIDIA shares declined after reporting earnings that beat consensus across every metric, marking the fourth consecutive post-earnings slide for the stock. The pattern signals a market that has already priced in NVIDIA’s dominance and is looking for evidence of the “what next.” Asian chipmakers rallied 5.5% on the back of NVIDIA’s AI demand outlook, suggesting the market continues to rotate into broader semiconductor beneficiaries (Intel, AMD, Micron, Qualcomm) rather than concentrating further in NVIDIA. Qualcomm surged 18% on the week.
Read More →The S&P 500 posted three consecutive losing sessions early in the week (falling to 7,353 on Monday) as the 30-year Treasury yield hit its highest level since 2007 and the 10-year touched a one-year high. The selloff was driven by fears that a prolonged U.S.-Iran conflict would keep oil prices elevated and put upward pressure on inflation. The market recovered fully by Friday as yields retreated and geopolitical de-escalation signals emerged. The episode illustrates how rate sensitivity remains the key vulnerability for tech and growth stocks in 2026.
Read More →Supply Chain & Commodities
CHIPS · MATERIALS · FREIGHTOn NVIDIA’s earnings call, CFO Colette Kress stated hyperscale capex is on track to exceed $1T in 2027, with agentic AI proliferating across all industries to drive spending to $3-4T annually by decade’s end. CEO Huang separately estimated capex is “at a trillion dollars and growing toward three to four” — well ahead of Needham’s consensus showing $1.03T in 2028. The revision trajectory is up-only: 2026 hyperscaler capex has been raised from $650B to $725B to $830B in just three months. US hyperscalers are now projected to spend 92% of operating cash flow on AI capex in 2026, up from 41% in 2023.
Read More →Semiconductor stocks sold off sharply on May 20 after news of a potential strike at Samsung and a stake sale by TSMC rattled global chip supply chains. Micron dropped 7.2% in a single session. The Samsung labor action threatens HBM3E and advanced DRAM production at a time when memory supply is already critically short. A simultaneous TSMC stake sale raised concerns about insider confidence, though analysts attributed it to routine portfolio rebalancing. The episode highlights how fragile the AI chip supply chain remains — concentrated in a handful of Asian fabs with no near-term capacity alternatives.
Read More →Total 2026 compute capex — including hyperscalers ($725B), Oracle ($50B), Apple ($13B), neoclouds ($60B), China ($80B), sovereign programs ($60B), and others ($48B) — is tracking approximately $1.04 trillion, making 2026 the first trillion-dollar year of compute capital expenditure in human history. Dell’Oro Research had expected this milestone by 2029 just a year ago, but pulled it forward to 2026 due to AI demand. The power grid, advanced packaging (CoWoS), and HBM remain the three binding physical constraints preventing even faster deployment.
Read More →