๐๐๐-๐ ๐๐๐๐ ๐๐จ๐ฎ๐ง๐ญ๐๐จ๐ฐ๐ง: ๐๐ ๐๐๐ฒ๐ฌ โณ
๐๐'๐ซ๐ ๐ญ๐ก๐ซ๐ข๐ฅ๐ฅ๐๐ ๐ญ๐จ ๐ฐ๐๐ฅ๐๐จ๐ฆ๐ @SebastianThrun ๐ญ๐จ ๐๐๐-๐ ๐๐๐๐!
Sebastian is the founder of Google X and co-founder of Udacity โ a pioneer of self-driving cars and physical AI, and a leading voice on how embodied intelligence will reshape the world.
๐๐ฉ๐๐๐ค๐ข๐ง๐ ๐จ๐ง ๐๐๐๐๐๐๐๐ ๐๐ โ the next frontier of embodied intelligence: self-driving, robotics, and machines that act in the physical world.
๐๐ ๐๐ญ ๐ ๐๐ซ๐จ๐ฌ๐ฌ๐ซ๐จ๐๐๐ฌ: ๐๐๐๐ก๐ง๐จ๐ฅ๐จ๐ ๐ฒ, ๐๐จ๐ฐ๐๐ซ, ๐๐๐ฉ๐ข๐ญ๐๐ฅ & ๐ญ๐ก๐ ๐๐๐ฑ๐ญ ๐๐ง๐ฏ๐๐ฌ๐ญ๐ฆ๐๐ง๐ญ ๐๐ฒ๐๐ฅ๐
๐๐๐๐ ๐๐๐๐. ๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐.
NEX-T Summit is an exclusive, invitation-only gathering of leading investors, visionary founders, business leaders, pioneering scientists, and top AI researchers. Through candid discussions, thought-provoking debates, and meaningful face-to-face connections, NEX-T explores how emerging technologies, global capital, and public policy are reshaping industries and creating the next generation of investment opportunities.
๐ Stanford Faculty Club, December 5, 2026
๐๏ธ Limited seats, registration subject to approval โ reserve your seat here: luma.com/NEX-T2026
#AI #NEXT2026 #Stanford #Investing #ArtificialIntelligence
๐๐๐-๐ ๐๐๐๐ ๐๐จ๐ฎ๐ง๐ญ๐๐จ๐ฐ๐ง: ๐๐ ๐๐๐ฒ๐ฌ โณ
๐๐'๐ซ๐ ๐ญ๐ก๐ซ๐ข๐ฅ๐ฅ๐๐ ๐ญ๐จ ๐ฐ๐๐ฅ๐๐จ๐ฆ๐ @FutureJurvetson ๐ญ๐จ ๐๐๐-๐ ๐๐๐๐!
Steve is Co-founder of Future Ventures and Draper Fisher Jurvetson โ one of Silicon Valley's most influential early-stage investors, backing SpaceX, Tesla, and Planet, and a leading voice on where AI capital goes next.
๐๐ฉ๐๐๐ค๐ข๐ง๐ ๐จ๐ง ๐๐ ๐๐๐๐๐๐๐๐๐ โ the next AI investment cycle, and where the next AI unicorns will come from.
๐๐ ๐๐ญ ๐ ๐๐ซ๐จ๐ฌ๐ฌ๐ซ๐จ๐๐๐ฌ: ๐๐๐๐ก๐ง๐จ๐ฅ๐จ๐ ๐ฒ, ๐๐จ๐ฐ๐๐ซ, ๐๐๐ฉ๐ข๐ญ๐๐ฅ & ๐ญ๐ก๐ ๐๐๐ฑ๐ญ ๐๐ง๐ฏ๐๐ฌ๐ญ๐ฆ๐๐ง๐ญ ๐๐ฒ๐๐ฅ๐
๐๐๐๐ ๐๐๐๐. ๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐.
NEX-T Summit is an exclusive, invitation-only gathering of leading investors, visionary founders, business leaders, pioneering scientists, and top AI researchers. Through candid discussions, thought-provoking debates, and meaningful face-to-face connections, NEX-T explores how emerging technologies, global capital, and public policy are reshaping industries and creating the next generation of investment opportunities.
๐ Stanford Faculty Club, December 5, 2026
๐๏ธ Limited seats, registration subject to approval โ reserve your seat here: luma.com/NEX-T2026
#AI #NEX-T2026 #Stanford #Investing #ArtificialIntelligence
Jensen Is Out. The Question Now Is Who Writes Nvidiaโs Checks.
Jensen Huang has resigned as Nvidiaโs president, CEO and board member, effective immediately, ending a 33-year run. CFO Colette Kress is interim CEO and stays CFO. Former Synopsys CEO Harvey Jones becomes Nvidiaโs first-ever board chair. The board will search inside and outside the company for a permanent CEO.
The stock fell as much as 13% this morning before paring losses to around 4%. From last Fridayโs level, Nvidia has lost as much as 19%, more than $855 billion of market value at the low.
Hereโs what changes, and what doesnโt.
Why the board could act this fast. Huang owns only about 3.5% of Nvidia and has no super-voting shares. Unlike founders at Meta or Alphabet, he never had structural control, so the board could force a change within days of the story breaking. Compare that with $META, which just created a new non-voting share class that further protects Mark Zuckerbergโs control. Governance structure decided this outcome as much as the scandal did.
What doesnโt change, at least not quickly: the order book. Nvidiaโs revenue is set by hyperscaler capex. Microsoft, Meta, Amazon and Alphabet all raised or reaffirmed AI spending in the past three weeks. Those purchase commitments and the CUDA software ecosystem donโt disappear with a CEO change. The board says operations, customer commitments and the product roadmap will continue without interruption. Kress has been CFO since 2013 and knows the customers and numbers better than anyone outside of Huang.
What could change: Nvidiaโs role as the AI industryโs bank. Under Huang, Nvidia didnโt just sell chips. It invested heavily across its own customer base: AI labs, neoclouds and startups that then buy Nvidia GPUs. Bernsteinโs point is the sharpest one: the CFO who signed off on many of those deals now decides how many more checks to write. A new or interim leader is more likely to be cautious than to expand that strategy.
That matters most for the companies that rely on Nvidiaโs support. Neocloud stocks fell today. As covered here last month, $CRWV and similar names live on the spread between AI returns and financing costs. If Nvidia becomes a more conservative investor, part of that financing ecosystem gets more expensive.
Earnings in five days. Nvidia reports on November 17, as scheduled. Kress will present results she would have presented anyway, but now as both CEO and CFO. Expect the numbers to be strong, since they reflect orders placed months ago. The real questions on the call will be about strategy, investment commitments, China and the CEO search. A strong quarter wonโt answer those.
The open risks.
Washington: Huang has agreed to testify before Congress, and the national-security investigation is ongoing. Its findings could affect export licenses and government relationships.
The search: no permanent CEO timeline yet. Every month without one is a month of strategic drift risk.
Competition: $AMD just guided to a big Q4 on its MI450 ramp. A leadership vacuum at Nvidia is the best opening AMD has had.
Bottom line: Nvidiaโs next quarter doesnโt depend on whoโs CEO. Its next decade might. The demand is locked in by customers who just raised their budgets. Whatโs now uncertain is strategy, and especially the investment engine that helped finance a big part of the AI build-out. Watch what Kress says about capital commitments on Tuesday. Thatโs where the post-Jensen Nvidia will show first.
The chips will ship. The checks are the question.
1
145
$NVDAโs Biggest Risk Isnโt the Scandal. Itโs Losing the Founder at the Worst Possible Time.
Nvidia fell about 4% on Monday, another 6% on Tuesday, and dropped again after hours Wednesday after the Wall Street Journal reported the board is preparing to oust founder and CEO Jensen Huang. According to the Journal, directors could meet as soon as Wednesday. Nvidia, which neared a $5.8 trillion market value in October, is now below $5 trillion.
The personal details are dominating the headlines. For investors, the question is narrower: what does Nvidia actually lose if Huang leaves?
What happened so far, in order:
Late October: reports surfaced about Huangโs ties to an executive at Megaspeed, a Singapore company that, per the New York Times, bought about $4.8 billion of Nvidia chips over two years, and whom US intelligence reportedly suspected of working for Chinese intelligence. Nvidia called the interactions โnot unusual.โ China dismissed the report.
Nov 4: three House committee chairs requested a closed-door briefing with Huang on Nov 18.
Nov 8: Huang admitted to staff that he had been unfaithful to his wife. The board opened an independent investigation into whether he was targeted by a foreign intelligence effort, saying it had no reason to believe secrets were compromised and expressing confidence in him.
Nov 11: the NYT reported a second relationship with a senior executive who reported to him. Hours later, the WSJ reported the board was preparing to remove him.
What Nvidia doesnโt lose: demand. Nvidiaโs revenue is set by the capex budgets of its biggest customers, and those just went up again. Microsoft, Meta, Amazon and Alphabet all signaled more AI infrastructure spending in the past three weeks. Those orders are placed for chips, not for a CEO. Supply contracts, product roadmaps and the CUDA software ecosystem donโt disappear in a leadership change.
What Nvidia could lose: three things that are hard to price.
The architect. Huang has run Nvidia since founding it in 1993 and personally drove its bet on accelerated computing and AI. Founders who set strategy, product cadence and culture for three decades are rarely replaced without disruption. No successor has been publicly named.
The relationships. Many of Nvidiaโs biggest partnerships, with hyperscalers, AI labs and governments, run through Huang personally. A sudden change at the top creates uncertainty for partners at a time when $AMD is winning major GPU deals with Meta and OpenAI.
Washington. This is now a national-security story. Nvidiaโs China exposure and export-control standing already make it politically sensitive. A congressional briefing request and an espionage-related probe add regulatory risk regardless of who leads the company.
Why the timing is so bad. Nvidia reports earnings on November 17. The congressional briefing request is for November 18. A company facing a possible CEO change, an open investigation and a politically charged hearing in the same week will struggle to keep the conversation on data center revenue, no matter how strong the numbers are.
What would calm this down. A clear, fast decision from the board, either way, and a credible succession plan. Markets can price a known outcome. What they canโt price is an open-ended leadership question. Leadership crises can also move fast and reverse: in 2023, OpenAIโs board fired Sam Altman and reinstated him within days.
Bottom line: Nvidiaโs business isnโt broken. AI demand is set by customers who just raised their budgets. But the company may be about to lose the founder who built it, in the middle of a national-security probe, days before earnings. The stock isnโt pricing lower demand. Itโs pricing uncertainty, and that wonโt clear until the board acts and the investigation reports.
The orders donโt depend on Jensen. The strategy and the relationships might.
๐๐ ๐ข๐ฌ ๐๐ญ ๐ ๐๐ซ๐จ๐ฌ๐ฌ๐ซ๐จ๐๐๐ฌ - ๐๐ง๐ ๐ญ๐ก๐ ๐ฉ๐๐จ๐ฉ๐ฅ๐ ๐๐๐๐ข๐๐ข๐ง๐ ๐ฐ๐ก๐๐ญ ๐๐จ๐ฆ๐๐ฌ ๐ง๐๐ฑ๐ญ ๐ฐ๐ข๐ฅ๐ฅ ๐๐ ๐ข๐ง ๐จ๐ง๐ ๐ซ๐จ๐จ๐ฆ ๐จ๐ง ๐๐๐๐๐ฆ๐๐๐ซ ๐.
๐๐๐-๐ ๐๐ฎ๐ฆ๐ข๐ญ ๐๐๐๐ is coming to Stanford University. 350+ AI researchers, founders, investors, and policymakers. One day. ๐
๐ข๐ฏ๐ ๐๐ข๐ ๐ช๐ฎ๐๐ฌ๐ญ๐ข๐จ๐ง๐ฌ:
1. After the Midterms, Who Will Shape the Future of AI Regulation โ and What Will It Mean for Markets?
2. Is the AI Boom Running Out of Steam โ or Is the Biggest Investment Wave Still Ahead?
3. Beyond LLMs, Where Will the Next AI Breakthrough Come From?
4. Can AI Transform Medicine โ and Create the Next Trillion-Dollar Industry?
5. Who Will Build the Next Generation of AI Giants?
๐
๐๐๐ญ๐ฎ๐ซ๐๐ ๐๐ฉ๐๐๐ค๐๐ซ๐ฌ
โข ๐๐๐๐๐ฌ๐ญ๐ข๐๐ง ๐๐ก๐ซ๐ฎ๐ง โ Founder of Google X and Co-founder of Udacity.
โข ๐๐ฎ๐๐ก๐ข๐ซ ๐๐ก๐๐ซ๐ฆ๐ โ Chairman of Rockefeller International; Founder and Chief Investment Officer of Breakout Capital.
โข ๐๐ญ๐๐ฏ๐ ๐๐ฎ๐ซ๐ฏ๐๐ญ๐ฌ๐จ๐ง โ Co-founder of Future Ventures and Draper Fisher Jurvetson. A longtime backer of Elon Musk's ventures.
โข ๐๐ซ๐จ๐. ๐๐ข๐๐ก๐๐๐ฅ ๐๐ง๐ฒ๐๐๐ซ โ Ascherman Professor of Genetics and Director of the Center for Genomics and Personalized Medicine at Stanford University School of Medicine.
โข ๐๐จ๐ฌ๐๐ฉ๐ก ๐. ๐๐ฎ, ๐๐, ๐๐ก๐ โ Professor of Medicine and Radiology at Stanford University, Director of the Stanford Cardiovascular Institute, and Co-founder of Greenstone Biosciences.
โข ๐๐ซ. ๐๐ฅ๐๐ฑ ๐๐๐๐ซ๐จ โ Distinguished Scientist & Head of AI Incubation, Zoom; Member, U.S. National Academy of Engineering. Previously Chief Scientist of Siri at Apple and a senior research leader at Microsoft.
โข ๐๐ซ๐จ๐. ๐๐ง๐ฎ ๐๐ซ๐๐๐๐จ๐ซ๐ โ Henry L. Moses Professor of Law and International Organization, Columbia Law School. An internationally renowned authority on AI governance, U.S. technology regulation, and global digital policy.
โข ๐๐ฎ๐ง ๐๐ข๐๐ง โ Chief AI Officer, GMI Cloud | Former Vice President of Generative AI Services, Oracle.
โข ๐๐๐ฏ๐ข๐ง ๐๐จ๐จ๐ฌ๐ โ Contributing Writer at The Atlantic and Co-host of Machine Gods.
Explore the full speaker lineup: luma.com/NEX-T2026
๐ ๐๐ญ๐๐ง๐๐จ๐ซ๐ ๐
๐๐๐ฎ๐ฅ๐ญ๐ฒ ๐๐ฅ๐ฎ๐, ๐๐๐๐๐ฆ๐๐๐ซ ๐, ๐๐๐๐
๐๏ธ ๐๐ข๐ฆ๐ข๐ญ๐๐ ๐ฌ๐๐๐ญ๐ฌ, ๐ซ๐๐ ๐ข๐ฌ๐ญ๐ซ๐๐ญ๐ข๐จ๐ง ๐ฌ๐ฎ๐๐ฃ๐๐๐ญ ๐ญ๐จ ๐๐ฉ๐ฉ๐ซ๐จ๐ฏ๐๐ฅ โ luma.com/NEX-T2026
โ๏ธ ๐
๐จ๐ซ ๐๐๐ฆ๐จ ๐๐ก๐จ๐ฐ ๐๐ฉ๐ฉ๐ฅ๐ข๐๐๐ญ๐ข๐จ๐ง๐ฌ, ๐ฌ๐ฉ๐จ๐ง๐ฌ๐จ๐ซ๐ฌ๐ก๐ข๐ฉ๐ฌ, ๐๐ง๐ ๐ฉ๐๐ซ๐ญ๐ง๐๐ซ๐ฌ๐ก๐ข๐ฉ๐ฌ: [email protected]
#AI #NEXT2026 #Stanford #Investing #ArtificialIntelligence #AI #Technology #SiliconValley
3
85
The Midterms Didnโt Change Who Writes the Laws. They Changed the Pressure on Oil.
Democrats are set to take control of both the House and the Senate in January, with flips in places few expected, including Texas. Politically, itโs a historic night. For investors, the question is simpler: what actually changes?
What doesnโt change: the veto. The president remains in office and keeps his veto. Overriding one takes two-thirds of both chambers. So sweeping new legislation from either side is unlikely. What does change is that Republicans can no longer pass tax and spending packages through reconciliation without Democratic votes. The realistic outcome for the next two years is gridlock, and markets have historically handled gridlock well. In the six months after past midterms, the S&P 500 has averaged a 12.4% gain, positive in all 13 cases measured. History isnโt a forecast, but itโs a useful baseline against panic.
What could change: the oil premium. Since the US and Israel struck Iran in February, oil has been the master variable for this market. It fed CPI, which fed rate-hike fears, which hit tech. It squeezed airline margins and drove the โBlack Thursdayโ selloff. Post-election commentary from both sides points to the same issues. Vice President Vance cited the war in Iran and the cost of living. Democratic leaders called the result a mandate on affordability.
A new Congress canโt end a war by itself. But it controls funding and oversight, and political pressure from both partiesโ voters now points in the same direction. If that pressure speeds up de-escalation, the war premium in crude could start to come out of the price. That would be bearish for energy names like $XLE, and supportive for everything oil has been weighing on: lower inflation, lower yields, higher tech multiples. The timeline is uncertain, and the new Congress isnโt seated until January. But this is the most important second-order effect of the night.
Sectors to watch.
Healthcare. The most legislatively sensitive sector. Democrats want to reverse cuts to Medicaid and healthcare spending, but any reversal still needs the presidentโs signature. Managed-care names with heavy Medicaid exposure may trade on expectations before any law passes. Drug-pricing pressure on pharma could return to the agenda. Note that healthcare stocks were already rallying ahead of the vote as Democratic odds rose.
Defense. Contracts already on the books arenโt affected by a change in Congress. The bigger swing factor is what happens with Iran, not who chairs the committees.
AI and Big Tech. More congressional scrutiny of AI and large tech companies is likely. Oversight hearings donโt change earnings next quarter, but they can add headline risk to the most crowded trade in the market.
Consumer. If affordability becomes the central policy fight, programs like SNAP and Medicaid directly affect lower-income spending and the retailers that depend on it.
How surprising was this? Not entirely. Prediction markets had priced a Democratic House as close to certain and gave Democrats better-than-even odds in the Senate. The scale of the result is the new information, not the direction.
Bottom line: for markets, this election is less about new laws and more about new pressure. Expect two years of gridlock on legislation, which markets usually tolerate well. The variable to watch is oil. If the political shift accelerates an end to the conflict, the biggest beneficiary may not be any sector on a policy list, but every asset the war premium has been holding down.
Gridlock on laws. Pressure on oil. Watch the barrel.
41
$SMCI Nearly Quadrupled Revenue. Read the Fine Print Before Celebrating.
Revenue of $19.9B, up 298% from $5.0B a year ago, crushing the ~$16.3B estimate. Adjusted EPS of $1.81 vs about $1.62. Net income of $1.59B, from $168M a year ago. Full-year fiscal 2027 guidance raised to $72โ82B, above the ~$69B consensus. The stock jumped more than 8% after hours.
Supermicro builds and assembles the server racks hyperscalers fill with $NVDA and $AMD chips. Every capex number from last weekโs earnings, from Microsoft to Meta to Amazon, ultimately flows to companies like this one. So the demand is real. But the quality of this quarter deserves a closer look.
1. Part of the beat is timing. Last quarter, revenue came in at $11.1B, near the low end of guidance. Management blamed customer delays: data centers whose power and cooling werenโt ready, with the revenue expected to land in later quarters. It landed. Revenue jumped about 80% in one quarter and beat Supermicroโs own $14.5โ15.5B guide by roughly 28%. Some of this quarterโs strength is last quarterโs delayed shipments.
2. The margin fell back to normal. Gross margin came in at 12.7%, down from 17.5% last quarter. That 17.5% was the outlier. Management said most of it came from a favorable customer and product mix. Over the past twelve months, gross margin averaged about 10.8%, down from 18% three years ago. Supermicro is a high-volume, low-margin assembler. It wins on scale, not pricing power.
3. The full-year guide implies a slower second half. Q1 revenue of $19.95B plus the Q2 guide midpoint of $20.5B gives about $40.5B for the first half. The full-year midpoint of $77B leaves about $36.5B for the back half, roughly $18.3B a quarter, below the first-half run rate. That could be lumpy deliveries or conservative guidance. Either way, the full-year number doesnโt describe a straight line up.
4. Cash and concentration remain the core risks. Operations used $434M of cash this quarter. Two quarters ago, the company burned $6.6B as inventory swelled past $11B and net debt hit $7.5B. Building racks means buying billions of dollars of GPUs and components before getting paid. Customer concentration is also extreme: earlier this fiscal year, one data center customer made up about 63% of quarterly revenue. When a few buyers drive most of the business, timing swings like the last two quarters are almost guaranteed.
Why the stock can still go up. Valuation. Even after its rally, Supermicro traded at roughly 11.5 times earnings in October, about half the S&P 500โs multiple. The market already prices in low margins, lumpy revenue, cash strain and past legal clouds. A quarter this big against expectations that low gives room for a pop.
Bottom line: a blowout quarter that confirms AI infrastructure spending is flowing through to the hardware layer. But itโs a timing-boosted print in a low-margin, cash-intensive, highly concentrated business. The demand is real. The question is whether Supermicro can turn it into steady margins and cash instead of a few big quarters at a time.
Huge revenue. Thin margins. Lumpy by design.
2
213
$AMDโs MI450 Ramp Is Real. So Is the Price of Winning It.
Revenue of $13.4B beat $13.1B. EPS of $2.09 beat $1.94. Data center revenue hit a record $8.55B, up 97% year over year and 27% in a single quarter, and now makes up 64% of AMDโs revenue. Three years ago it was about 28%. The stock jumped more than 6% after hours.
But the quarter isnโt the headline. The guide is.
Q4 guidance: about $17B. Against $13.4B this quarter, thatโs roughly 27% sequential growth, and about 66% above last yearโs Q4. Itโs well ahead of the ~$16.2B estimate. This is the MI450 ramp Lisa Su has been promising: AMDโs new AI GPUs, shipping in Helios racks to Meta, OpenAI, Microsoft, Oracle and Anthropic, plus a $12B HPE deal for Vultr. The question for a year was whether AMD could turn AI design wins into AI revenue at scale. A $17B guide is the answer.
The margin held. MI450 carries lower gross margins than AMDโs average, so a big ramp could have dragged the company margin down. Instead, Q4 gross margin guidance is about 55%, close to Q3โs 56%. Growing revenue 27% in a quarter without sacrificing margin is the best version of this ramp.
Now the cost: the warrants. To win its two biggest AI deals, AMD paid partly in its own stock. OpenAI and Meta each hold warrants to buy up to 160 million AMD shares at one cent per share, roughly 10% of the company each. They vest as each customer buys up to 6 gigawatts of AMD GPUs and as AMDโs stock hits targets rising to $600. In effect, AMD is giving customers equity to secure purchase commitments.
Two things follow:
Dilution. If both warrants vest fully, existing shareholders could be diluted by around 20% combined.
A charge that grows with success. AMD remeasures the warrants every quarter, and this quarter that cost $1.097B. Thatโs why GAAP net income was $2.43B while adjusted net income was $3.47B. The higher the stock goes, the bigger the charge. Itโs the same accounting dynamic behind the giant non-cash losses in Anthropicโs IPO filing.
Supporters call these deals smart customer lock-in. Skeptics call them circular financing. The truth is both: they guarantee demand, and shareholders fund part of it.
The weak spot: PCs. Client and Gaming revenue fell 8% to $3.70B. Last quarter, client grew 23%. This is the same memory squeeze that crushed $INTC last week: memory prices are making PCs too expensive, and shipments are falling. The difference is AMD has a data center business big enough to absorb the hit. Intel doesnโt.
Remember last quarter. In August, AMD also beat and guided above consensus, and the stock still fell about 9% after hours. Expectations are high, and the bar keeps rising.
Bottom line: the MI450 ramp is real, the Q4 guide is a big step up, and margins are holding. That puts AMD firmly in the AI accelerator race, not just alongside $NVDA in slides. The price is equity: up to roughly a fifth of the company promised to two customers, and a GAAP charge that grows as the stock does. Data center is carrying the company while PCs shrink, and thatโs now the whole investment case.
The ramp is real. The warrants are the toll.
4
563
$AAPL Beat. The Memory Bill Is Why the Stock Isnโt Celebrating.
Revenue of $114.5B beat $113.4B and topped the high end of Appleโs own guidance. EPS of $2.10 beat $1.99. iPhone hit a September-quarter record at $57.6B, up 17%. Mac crushed estimates at $10.8B. Fiscal 2026 was Appleโs best year ever: $479B in revenue, up 15%, with $86.4B spent on buybacks.
And the stock is drifting lower. Three numbers explain it.
1. The margin is shrinking underneath. Gross margin came in at 47.9%, the top of guidance. But about 1 point of that came from one-time tariff refunds. Strip those out and underlying margin was roughly 46.9%. Last quarter, underlying margin was about 48.1%. Thatโs a 1.2-point drop in one quarter, and Apple has said memory costs explain all of the decline. IDC estimates memory costs are up around 350% from last year. Tim Cook called it a โ100-year floodโ on his final call. It hasnโt receded.
This is the same story playing out across the sector: memory makers like $MU and Sandisk post record margins, while everyone who buys memory pays for it. It crushed $INTCโs PC business, pushed Amazonโs capex higher, and now itโs eating into Appleโs gross margin.
2. Services keeps slowing. Services grew 10% to $31.6B, down from 12% last quarter. Services is Appleโs highest-margin business, with gross margin around 75%, so itโs supposed to be the cushion that offsets hardware cost pressure. A cushion growing slower is a thinner cushion, right when hardware margins need it most.
3. China missed. Greater China revenue came in at $17.5B, against $18.6B expected. Itโs still up 21% year over year, but missing by about $1B in the region with the most competition matters.
The real test is the December quarter. This was John Ternusโs first quarter as CEO, and his first guide matters more than this print. The questions:
Margins: Bernstein estimates December-quarter iPhone gross margin could fall to around 40%, as cheaper memory inventory runs out and new memory costs more.
Volume: Loop Capital just downgraded Apple, modeling 79 million iPhones vs the Streetโs 85 million.
iPhone Duo: the new phone starting at $1,999 went on sale October 23. How many Apple can build, and how buyers respond to that price, lands entirely in the next quarter.
Appleโs answer to memory costs has been raising prices. That works as long as demand holds. The December quarter is when that bet gets tested at the biggest scale.
Bottom line: a record year and a clean beat, but the margin is quietly being squeezed by memory, and Services is slowing just when itโs needed most. The stock is reacting to where the margin is heading, not where it landed. The holiday-quarter margin guide on tonightโs call is the real report.
Record revenue. Thinner margin. Memory sends the bill.
2
398
$PLTR Is Growing Faster Than Ever. Two Numbers Say Watch the Next Turn.
Revenue of $2.31B, up 96%, beating $2.18B. EPS of $0.44 vs $0.41. US commercial revenue up 148% to $984M. Full-year guidance raised to about $8.53B, roughly 91% growth, up from 82% guided last quarter. Alex Karp opened the shareholder letter by saying few companies in postwar America have grown this fast. On the numbers, he has a point.
Revenue growth has accelerated for eight straight quarters, from 36% in late 2024 to 96% now. At a $9B run rate, thatโs almost unheard of.
So the question isnโt whether this was a great quarter. Itโs what comes next. Two numbers matter more than the headline.
1. The backlog grew much more slowly. US commercial remaining deal value is contracted revenue that hasnโt been recognized yet. Itโs the best leading indicator of where commercial revenue is heading. Last quarter it grew 27% sequentially to $6.24B, up 124% year over year. This quarter it reached $6.5B: up just ~4% sequentially and 79% year over year. Revenue accelerated, but the pipeline feeding future revenue slowed sharply. One quarter isnโt a trend, since large deals land unevenly. But itโs the first real cooling in the metric bulls rely on.
2. The Q4 guide implies the streak ends. Q4 guidance of about $2.65B compares with roughly $1.41B a year ago, implying ~88% growth. That would be the first deceleration in two years. The caveat: Palantir guides conservatively. Q3 guidance was $2.16B and actual revenue came in about 7% higher. A similar beat in Q4 would put growth near 100%. Still, the companyโs own math is the first sign the acceleration curve may be flattening.
Why that matters at this price. After Q2, the stock jumped 29% in a day and traded at roughly 48 times 2026 revenue and 85 times adjusted free cash flow. Valuations like that price in years of extreme growth. At that level, it isnโt enough to be growing fast. The rate has to keep rising. Thatโs why Palantir has fallen after strong reports before, including a 6.9% drop after its Q1 beat.
Whatโs not in question: business quality. Last quarter Palantir posted a 62% adjusted operating margin, a Rule of 40 score of 155%, net dollar retention of 157%, and $9.2B of cash with no debt. US commercial growing nearly 150% means AI deployments are moving from pilots into production. Palantir isnโt a story stock. Itโs a highly profitable company growing at startup speed.
Bottom line: an extraordinary quarter with no real flaws in the reported numbers. The risk is all in the forward indicators. If RDV re-accelerates and Q4 beats its guide like Q3 did, the eight-quarter streak continues and the multiple can hold. If the pipeline keeps slowing, the market will start pricing the turn before it shows up in revenue.
The growth is real. The slope is the trade.
2
316
Why $SNDK Fell After a Double Beat: The Market Trades the Slope, Not the Level
Revenue of $11.03B beat $10.64B. Adjusted EPS of $49.51 beat $46.22. Gross margin hit a record 85.1%. Revenue is up 378% from a year ago, net income went from $112 million to $8.1 billion, and Q2 guidance is above consensus too. Plus an 8-for-1 stock split.
And the stock is red. The same thing happened in August. Hereโs why.
Growth is still huge, but the slope is flattening. Look at sequential revenue growth:
Fiscal Q4 2026: +51% ($5.95B โ $8.97B)
Fiscal Q1 2027: +23% ($8.97B โ $11.03B)
Fiscal Q2 2027 guide: ~+12% (midpoint $12.4B)
Growth is roughly halving every quarter. In a cyclical industry like memory, stocks donโt peak when earnings peak. They peak when the rate of improvement peaks. A decelerating slope is exactly what investors who lived through past memory cycles are trained to sell.
Margins have hit a ceiling. Gross margin went from 78.4% to 84.6% to 85.1%, and guidance is 84.5โ85.5%. Thereโs not much room above 85%. When margins stop expanding, earnings growth has to come entirely from volume and price, and price is whatโs slowing.
NAND price hikes are cooling. TrendForce estimated NAND contract prices rose 70โ75% in the spring quarter, then projected 10โ15% for the quarter after. Goldmanโs view is that hikes continue through Q4 2026, then prices start falling in 2027 as new supply comes online. Thatโs the bear case in one sentence: the best pricing may be right now.
The bull case is contracts. Unlike past memory cycles, this one is built on long-term agreements. Last quarter Sandisk reported $59.8B of remaining performance obligations, or $91.1B including deals signed after quarter end, with about half of fiscal 2027 bit shipments and two-thirds of fiscal 2028 already committed. Data center revenue jumped 61% sequentially to $4.80B, and thatโs where the contracts sit. Contracted volume at locked-in terms is the reason this cycle shouldnโt collapse the way 2019 or 2023 did.
The warning light in the report: consumer. Consumer revenue fell 39% year over year to $397M. High memory prices are crushing demand for consumer storage, the same dynamic that drove PC shipments down 20% and sank $INTC yesterday. At these prices, memory makers are winning data centers and losing everyone else.
About the split. An 8-for-1 split changes nothing about the value of the company. It makes shares cheaper per unit and easier to buy for employees and retail investors. Itโs a signal of confidence, not a fundamental catalyst.
Bottom line: this was a great quarter, and the market sold it anyway because memory stocks trade on direction, not level. Growth is decelerating, margins have plateaued, and pricing momentum is fading. The counterweight is a contract book unlike anything in past cycles. The real test isnโt this quarter. Itโs whether 2027 contract prices hold when new supply arrives.
Record earnings. Flattening slope. In memory, the slope wins.
2
589












