• Consider NVIDIA (NVDA) for exposure to sustained AI-infrastructure demand, while recognizing that power and construction constraints may delay deployments for the next couple of years.
• Watch Super Micro Computer (SMCI): preliminary revenue of $19.8 billion topped the $15 billion expectation, but confirm final results and guidance before acting.
• Alphabet (GOOGL) may be worth watching below $300, as the host suggested, but reassess after earnings for signs of worsening advertising trends.
• Treat SpaceX’s growth outlook as uncertain and wait for a significant pullback; planned $40 billion in Q4 CapEx, debt, and potential dilution are key risks.

SpaceX (no ticker stated)

• The host described SpaceX as the key catalyst for the AI trade after a strong earnings report: $14.6 billion revenue, up 161% year over year; $0.35 EPS versus a $0.16 expectation; $4.4 billion net income; and $13.1 billion operating cash flow.
• SpaceX reportedly had 1.7 gigawatts of compute capacity online. Elon Musk said the company expected about 13 gigawatts by the end of 2027, with a possibility of exceeding 20 gigawatts.
  • Musk also gave an ambitious end-of-2027 ARR range of $500 billion to $1 trillion. The host emphasized that these projections are uncertain.
• The host highlighted a 5% operating margin in the data-center business and said it could have been higher without a $1.2 billion stock-based compensation charge.
• Musk said demand for compute was exceeding available supply, and that short-term contracts were renewing at higher prices per watt.
• The host cited analyst price targets ranging from $205 to $500, including a JPMorgan target of $500. These are analyst estimates, not guarantees.
• Risks discussed included a planned $40 billion in Q4 CapEx, versus an $18 billion Street expectation; the likelihood of raising additional capital; possible dilution; $40 billion in debt; a large share unlock; and uncertainty around a potential Tesla merger. Musk said the companies’ special committees had not agreed to merger terms.

Takeaways

• The earnings report strengthened the case for sustained compute demand, but the growth projections require substantial investment and execution. Treat the 2027 capacity and ARR figures as management’s targets, not established outcomes.
• The host remained cautious about the valuation and said they would consider SpaceX more seriously after a significant pullback. Any Tesla merger could add complexity and dilution risk.

CoreWeave (CRWV), Nebius (NBIS), and other neoclouds

• SpaceX’s results were presented as evidence that demand for AI compute is strong, potentially supporting neocloud providers such as CoreWeave and Nebius.
• Despite that demand argument, CoreWeave and Nebius shares struggled during the session. The host said investors may prefer SpaceX because of its scale, reported profitability, access to capital, and ability to bring capacity online quickly.
• CoreWeave’s financing was a major concern. The host cited a market yield of roughly 15.2%–15.3% for its bonds, while clarifying that this was not the coupon CoreWeave currently pays.
  • Brad Gerstner argued that lenders may see CoreWeave as riskier because its GPU infrastructure is less straightforward to finance than NVIDIA systems, particularly if custom chips are involved.
• The host said CoreWeave was not expected to reach GAAP profitability until 2029, and Nebius until 2030. Nebius had previously raised capital at higher share prices, which the host viewed as a relative financing advantage.
• IREN, Cipher Mining (CIFR), TeraWulf (WULF), and Applied Digital (APLD) were also discussed as data-center or compute-related companies. IREN announced a 300-megawatt site in Spain, though its share-price gains faded during the session.
• The host’s view was that SpaceX validates the broader compute market, but does not remove company-specific financing and execution risks.

Takeaways

• Neocloud exposure offers potential upside if compute demand remains strong, but financing costs, capital needs, and the ability to bring facilities online are central risks.
• The transcript does not establish that weakness in CoreWeave or Nebius is a buying opportunity. The host said buying the dip depends on confidence in the companies’ long-term execution.

NVIDIA (NVDA) and AI infrastructure

• NVIDIA was described as a potential beneficiary of SpaceX’s planned data-center buildout and continuing demand for GPUs.
• At GTC, Jensen Huang said physical constraints—including power, land, construction, wafers, memory, and networking components—were limiting deployment. He said bottlenecks may persist for a couple of years.
• NVIDIA announced DSX Pod, a modular AI-factory design that the host said could reduce deployment time from six months to as little as 14 days.
• The host also cited NVIDIA’s prior growth outlook of roughly 70%, with Huang suggesting growth could be higher if supply allowed. These figures were discussed as management commentary, not a new forecast in the episode.
• Huang said NVIDIA GPUs could remain useful for seven to eight years, supported by software updates and the CUDA ecosystem.
• Risks discussed included power availability and other physical limits on deploying compute.

Takeaways

• The discussion supports a bullish view of AI infrastructure demand, while highlighting that physical deployment constraints may limit how quickly revenue can be realized.
• Watch for evidence that customers can install and use the equipment they purchase; hardware demand alone does not guarantee immediate deployment or returns.

Super Micro Computer (SMCI) and Dell (DELL)

• SMCI reported preliminary revenue of $19.8 billion, versus a Street expectation of $15 billion, according to the host. Its shares rose sharply in pre-market trading.
• The host linked SMCI’s results to its work with SpaceX and the broader AI-server market.
• Dell also rose, with the host suggesting SMCI’s preliminary numbers supported the wider server and AI-infrastructure trade.
• The host noted that SMCI has faced concerns related to China, but said the preliminary revenue figures were encouraging.

Takeaways

• The preliminary revenue result is a positive signal for AI-server demand, but it is not a substitute for final results and guidance.
• The transcript does not resolve SMCI’s company-specific risks or establish that its valuation is attractive.

Micron (MU), Amkor (AMKR), and semiconductor suppliers

• Micron was discussed as a memory supplier benefiting from the AI buildout, but its shares repeatedly gave back early gains. The host said investors may be concerned about how long current supply shortages and pricing power will last.
• Jensen Huang’s comments that some bottlenecks could ease over the next few years added to questions about the durability of suppliers’ unusually strong margins.
• Amkor announced an expanded partnership with NVIDIA, including plans to increase advanced-packaging capacity in South Korea by more than 300% by the end of 2028.
• TSMC (TSM), Intel (INTC), Broadcom (AVGO), Marvell (MRVL), Qualcomm (QCOM), and Arm (ARM) were mentioned as parts of the semiconductor and AI supply chain.
  • The host noted that Intel and TSMC could potentially benefit if SpaceX builds out more chip capacity, but did not cite a specific confirmed contract.
• The host said supply constraints in areas such as memory, packaging, and photonics could persist, while also noting that eventual supply growth could weaken pricing power.

Takeaways

• AI-related demand may support semiconductor suppliers, but the key question raised was how long supply shortages and elevated margins will last.
• Amkor’s announced capacity expansion is a concrete development; for the other suppliers, the transcript mostly discussed potential exposure rather than confirmed new business.

Meta Platforms (META)

• Meta was due to report earnings after the market close. The host expected discussion of Muse, compute capacity, and capital spending.
• Muse had reportedly reached 11 million downloads. The host was looking for management to discuss user growth and the compute required to support the product.
• Meta’s stock was also affected by a report that a security flaw exposed 6,800 sharing links to Muse-generated content. The host said Meta had paused the affected functionality while reviewing the issue.
• The host mentioned a potential $10 billion legal expense as a possible source of pressure on the stock.
• Advertising trends were another uncertainty: the host noted a sell-side report describing slower trends and softer returns on investment for Google, and said similar trends could matter for Meta.

Takeaways

• Meta’s investment case in the discussion depended on whether Muse and other AI features can drive usage and justify higher compute spending.
• Before treating AI growth as a positive for the shares, consider the specific risks discussed: the potential legal expense, advertising conditions, and the reported product-security issue.

Alphabet (GOOGL)

• Google’s stock fell during the session, which the host attributed partly to concerns about OpenAI expanding advertising to more countries and a sell-side report citing slower advertising trends.
• The host argued that Google’s cloud business could still benefit from strong demand for compute. Musk had said hyperscalers were seeking more capacity.
• CEO Sundar Pichai said Google prioritizes having enough compute for frontier AI work, while planning capacity for its other businesses as well. He noted that current compute availability reflects investment decisions made years earlier.
• Google’s robotics work was also discussed, though the host said the showcased cornhole demonstration was early-stage and not evidence of near-term commercial scale.
• The host said they would be interested in buying Google below $300, provided earnings did not reveal a serious deterioration in the business.

Takeaways

• The discussion presented competing forces: AI and cloud demand on one side, advertising concerns and capital allocation on the other.
• The host’s below-$300 comment was a personal view, not a formal target. The earnings report and advertising commentary were identified as important checks on that view.

Amazon (AMZN) and Oracle (ORCL)

• Musk’s comments about hyperscalers seeking compute were viewed by the host as evidence of continued demand that could benefit cloud providers, including Amazon and Google.
• Amazon was also discussed in connection with an article about GPUs waiting to be brought online because of power and construction constraints. The host argued that deployment delays do not necessarily mean customers do not want the hardware.
• Oracle was mentioned among the AI infrastructure names that had been under pressure. NVIDIA’s modular DSX Pod announcement was presented as a possible way to shorten deployment timelines.
• Both stocks were weak at points during the session, despite the broader demand argument.

Takeaways

• Strong customer demand is not the same as immediate revenue: power, land, and construction can delay deployment.
• For both companies, the transcript points to a need to distinguish committed demand from the pace at which infrastructure can actually be brought online.

Tesla (TSLA)

• Tesla was due to report earnings after the close. The host said robotics and autonomous driving were important parts of the longer-term story.
• Musk said a SpaceX–Tesla merger could have strategic benefits, but also said the companies’ independent committees had not agreed on terms and there was no guarantee of a deal.
• The host emphasized that a merger could require substantial stock issuance or other financing, creating potential dilution for shareholders.
• Huang discussed autonomous vehicles as a large opportunity involving multiple vehicle brands and ride-hailing platforms, rather than a market necessarily dominated by one company.
• The host cautioned that humanoid robotics may take time to become commercially significant.

Takeaways

• Tesla’s earnings and commentary on autonomy, robotics, and any potential merger were identified as key catalysts.
• The discussion did not establish that a merger will happen. Investors should account for uncertainty around terms, dilution, and the timeline for robotics adoption.

Netflix (NFLX)

• Netflix’s earnings were described as weak, with misses on multiple metrics and a lower full-year EPS outlook.
• The company reduced its 2026 advertising revenue outlook from $3 billion to $2.7 billion, according to the host.
• The host said the stock fell about 7% after the report and questioned the durability of a content-driven subscription model when content momentum weakens.
• The host estimated a fair value of about $70 using a personal analysis prompt, and said $50–$55 would be more interesting for a long-term investment. The host also said Netflix could remain a strong company over time if it continues compounding EPS at roughly 20%.

Takeaways

• The discussion was cautious near the then-current share price because the earnings report did not provide a clear growth catalyst.
• The valuation figures and EPS-growth view were the host’s estimates. The main issues to monitor are advertising growth, content performance, and whether earnings growth can meet expectations.

Bitcoin (BTC) and Ethereum (ETH)

• The host said Bitcoin had been tracking the AI trade and could benefit if technology shares rallied.
• Bitcoin traded around $83,000–$85,000 during the discussion. Ethereum approached $2,500.
• The host said a move toward $95,000 for Bitcoin could make investors more confident that crypto had recovered, but did not give a specific recommendation.
• The transcript also discussed tokenization of financial assets as a longer-term potential use case for blockchain technology.

Takeaways

• The discussion framed crypto as sensitive to broader market sentiment, particularly movements in technology stocks.
• Tokenization was presented as a potential long-term theme, but the transcript did not identify a specific project or timeline that would make it investable now.

Zeta Global (ZETA) and digital advertising

• The host highlighted Zeta’s response to Salesforce’s purchase of Braze: Zeta compared its own growth and cash-flow measures with Braze’s and argued that it was undervalued relative to the deal.
• The host viewed the statement as a sign of confidence ahead of Zeta’s next earnings report.
• Salesforce (CRM) was mentioned in connection with the Braze deal, while Reddit (RDDT) and other digital advertising names were discussed as part of a broader advertising market.
• The host said Zeta had already risen significantly for the year, so any further upside would depend on earnings and guidance.

Takeaways

• Zeta’s comparison with Braze is management’s positioning, not proof that the market undervalues the stock.
• The next earnings report was identified as the key test of whether revenue growth and cash-flow expectations support the host’s positive view.

Bloom Energy (BE), Vistra (VST), and Constellation Energy (CEG)

• The host described Bloom Energy as a higher-growth energy opportunity tied to fuel cells and data-center power demand, and said it was working with Brookfield and Oracle.
• Vistra and Constellation Energy were also mentioned as energy companies receiving interest amid AI-related power needs.
• Jensen Huang’s comments reinforced power as a major physical constraint for AI data centers.
• Asked to choose between Bloom and Vistra, the host said Bloom for growth and Vistra for a greater valuation margin of safety.

Takeaways

• The transcript links power demand to AI infrastructure, but the potential opportunity depends on companies actually winning projects and delivering capacity.
• The host’s distinction was growth versus valuation comfort; it was not a claim that either stock is risk-free.

Grab (GRAB) and Uber (UBER)

• Taiwan blocked Grab’s planned Foodpanda acquisition, which the host said involved about $600 million.
• The host viewed the block as potentially positive for Grab’s near-term finances because the company would retain the money and could focus on its existing markets. The deal would otherwise have supported expansion into another country.
• The transcript also cited oil prices and currency movements as pressures on ride-hailing and other companies with international operations.
• Uber was mentioned as another company exposed to oil and currency conditions; no specific company news was cited during the discussion.

Takeaways

• For Grab, the blocked acquisition removes a potential expansion opportunity but also avoids the associated spending and execution demands.
• The host’s favorable interpretation is a possibility, not a confirmed improvement in Grab’s operating outlook. Monitor local growth and the effect of oil and currency movements.

Rocket Lab (RKLB), AST SpaceMobile (ASTS), and space-related equities

• Rocket Lab and AST SpaceMobile were mentioned as space-sector names that received some attention alongside SpaceX.
• The host noted that Rocket Lab had fallen substantially from prior highs and said delays to its Neutron rocket program had weighed on the stock.
• SpaceX’s earnings were described as a possible source of sympathy interest in other space companies, but the transcript did not identify a direct business benefit for them.

Takeaways

• A rally in SpaceX does not by itself validate other space companies. The transcript points to company-specific execution and program timelines—particularly Rocket Lab’s Neutron schedule—as important considerations.

Other companies mentioned

• Palantir (PLTR) was discussed as a potential beneficiary of on-premise AI systems and an NVIDIA partnership reference. The host acknowledged its valuation was high but argued that strong growth and profitability supported it.
• Robinhood (HOOD) was mentioned as a potential beneficiary of renewed crypto strength.
• SoFi (SOFI) was discussed favorably by one speaker, who cited growth, leadership, and insider buying; no new financial results were provided.
• Capital One (COF) was cited as reporting a resilient consumer and credit quality, though the host noted that strong spending could be concentrated among wealthier consumers.
• IBM (IBM), Texas Instruments (TXN), Lam Research (LRCX), SAP, and Southwest Airlines (LUV) were noted as upcoming or relevant earnings reports. The host said IBM’s prior commentary about customer spending and software budgets could remain a concern.
• D-Wave (QBTS), Moderna (MRNA), CrowdStrike (CRWD), ServiceNow (NOW), Arm (ARM), and several other stocks were mentioned in market commentary without detailed new investment analysis.

Takeaways

• The transcript offered limited company-specific evidence for these names. Their mentions should not be treated as recommendations.
• Earnings reports, valuation, and the ability to convert AI or consumer demand into sustained financial results were recurring issues across the stocks discussed.