The week handed us two of the cleanest AI infrastructure data points of the year, Google Cloud growing 82% and Intel posting its strongest revenue growth in fifteen years, and the tape sold both of them. That is not a thesis break. That is a market in the grip of two competing anxieties: a Hormuz-driven oil shock that won't quit and a capex-sticker-shock reflex that keeps confusing "spending more on the winning trade" with "spending too much." The fund's read: the AI infrastructure thesis just had its best earnings week of the cycle. The stock charts don't show it yet. They will.
The Market Punished the Right Answer. That Is the Setup.
"The AI revenue thesis is no longer a projection, it printed in the 10-Q. The selloff is a valuation debate, not a fundamental one."
Alphabet reported Q2 revenue of $119.8 billion, up 24% year-over-year, with Google Cloud accelerating to $24.8 billion, an 82% gain that nearly doubled Q1's already elevated 63% growth pace. This is not a rounding error or a favorable comparison. This is the AI monetization thesis printing at a velocity that exceeds even the most constructive estimates entering the quarter. The cloud backlog entering Q3 stands above $460 billion. The enterprise AI demand signal is unambiguous.
Alphabet beat Q2 estimates on every headline metric, and the stock still fell from around $374 to $320 in two trading days, with AI capex fears, a new EU fine, and $500 billion in new spending commitments cited as the drivers. The market's reaction is rational at the margin and wrong at the thesis level. Alphabet is not spending $180–190 billion on a guess. It is spending it on a Google Cloud backlog it cannot fulfill fast enough.
The structural read: when a company generates 82% cloud growth and the stock drops 15%, the market is not saying the AI trade is broken. It is saying the AI trade's valuation premium no longer provides a cushion for any disappointment, real or perceived. Alphabet raised full-year 2026 capex guidance to $180 to $190 billion and signaled that 2027 spending will "significantly increase." That spooked the margin-focused crowd. Our view: any capex spend backed by a $460-billion committed backlog is not overreach, it is the rational response to supply being the binding constraint.
Pillar by Pillar: What Actually Moved and What It Means
AI Chips / Hardware: Intel's 18A Moment Changes the Foundry Calculus
Intel delivered Q2 revenue of $16.1 billion, exceeding guidance by $1.8 billion, with non-GAAP EPS of $0.42 against a $0.20 estimate and gross margin 280 basis points above guidance. AI-driven businesses grew over 70% year-over-year. This is not a turnaround tease, it is a confirmation. Intel's revenue jumped 25% year-over-year, the most robust growth for any period since the third quarter of 2011.
Intel's 18A production node exceeded internal volume targets by approximately 25% in Q2, with output up over 50% quarter-over-quarter, and CEO Lip-Bu Tan highlighted strong execution on 18A and 18A-P nodes with yields tracking ahead of schedule. The foundry story, the one the Street has been skeptical of for three years, is now producing real numbers. The Data Center AI segment posted record year-over-year server growth of 59%, driven by Xeon 6 demand.
For the broader AI chip complex, NVDA, AMD, AVGO, the week's message is consistent: the infrastructure demand is accelerating, not decelerating. Microsoft confirmed an expanded long-term partnership with AMD deploying next-generation AMD Instinct and EPYC processors (Jul 20), a quiet but consequential vote of confidence in the multi-vendor AI accelerator thesis. Intel fell 7.89% on Friday in sympathy with the broader chip selloff, even with a blowout print, this is the sell-the-news dynamic that created the AVGO entry thesis six weeks ago and is creating a similar setup in INTC now.
AI Software / Platforms: Google Cloud's 82% Growth Is the META Capex Validation
Google Services revenues rose 15% to $94.5 billion, led by 17% growth in Search and 13% in YouTube ads. The Street was focused on whether AI Overviews would cannibalize search monetization. The Q2 data answers that: search is still growing double-digits and Cloud is growing at 82%. That is not a cannibalization story, that is a monetization expansion story. META reports next week. Given what Google Cloud just printed, Meta's AI ad platform thesis, which rests on the same enterprise AI monetization current, deserves elevated conviction heading into the print.
It is shaping up to be a strong earnings season for the S&P 500 broadly. FactSet data shows the year-over-year S&P 500 earnings growth rate for Q2 tracking at 24.7%, well above the five-year average of 16.4% and the ten-year average of 10.3%. The context matters: the AI infrastructure pillar is not leading a weak earnings season, it is leading the strongest earnings season in years.
Nuclear / Power Infrastructure: Oil at $100 Is a Structural Double-Edged Signal
Oil rose to a six-week high midweek as the U.S. and Iran traded further strikes, with Brent crude touching $93.40 per barrel, levels not seen since early June, as a brief détente between Washington and Tehran ended earlier this month, restricting tanker movement through the key Strait of Hormuz. The international benchmark touched $100 per barrel on the week before pulling back Friday as Pakistan-mediated diplomacy rumors circulated. Brent fell roughly 4% Friday to trade below $96.
The national average price of gasoline climbed back above $4 a gallon as fighting between the U.S. and Iran intensified, disrupting one of the world's most important oil shipping routes, with AAA reporting $4.003 per gallon on Monday, up from $3.872 a week prior and $3.141 a year ago. This is structurally bullish for nuclear and domestic power infrastructure. Every dollar Brent adds above $90 widens the competitive moat for zero-fuel-cost baseload generation, exactly what VST and CEG supply to AI data center operators locked into long-term power purchase agreements.
The nearly five-month war has depleted global stockpiles and stoked inflation worldwide. The Fed cannot cut into a Hormuz-driven oil shock. That means higher-for-longer rates persist, which compresses the multiple on long-duration growth equities but strengthens the cash-flow case for power infrastructure assets with contracted revenue. The power pillar is the one place in the portfolio where geopolitical chaos is a tailwind, not a headwind.
Digital Finance Rails: The GENIUS Act Missed Its Own Deadline
U.S. agencies missed the GENIUS Act's July 18 statutory deadline to finalize stablecoin rules, leaving issuers without a federal framework ahead of the January 2027 enforcement date. This is not fatal to the thesis, the January enforcement date remains intact, and the regulatory arc is still clearly heading toward institutionalization of dollar-denominated stablecoins. But the delay is a tactical negative for any name dependent on near-term regulatory clarity as a catalyst. We are watching this pillar carefully. The stablecoin legislative framework remains the structural thesis; the slippage in rule-finalization timing is a catalyst delay, not a thesis break.
Bitcoin was trading near $64,784 as of Monday's open, largely range-bound amid the risk-off oil shock. Crypto broadly tracked the tech retreat this week, which is consistent with the "risk-off correlates everything" dynamic we see at Brent above $90. The Digital Rails pillar thesis, stablecoin adoption as institutional payment infrastructure, is intact. The timeline is simply shifting toward Q4 as the catalyst window.
Next Week: The Fed, the Grand Finale, and the $725 Billion Checkpoint
Next week is the most consequential seven-day window of the quarter. Three catalysts dominate:
1. FOMC: July 28–29. The FOMC is in its pre-meeting blackout ahead of the July 28–29 meeting, so there have been no Fed voices to move the tape this week. The market enters that meeting pricing no rate change, but the language around oil-driven inflation risks and the pace of any future easing is the lever that moves long-duration tech. With Brent having touched $100 this week and gasoline nationally above $4, Chair Powell's press conference tone on inflation risks is the swing variable. A hawkish-lean surprise would be a near-term headwind for the AI software and chip names that already repriced this week.
2. Hyperscaler Grand Finale: META, MSFT, AMZN, AAPL. The expectation heading into the meeting is that the Fed keeps rates on hold, and the July 29 statement will likely reveal what data the FOMC is monitoring to determine the path forward. But the earnings prints matter more for the AI thesis. Meta, Microsoft, Amazon, and Apple all report in the next five trading days. The collective capex guidance from those four, $725 billion combined for 2026, up 77% from $410 billion in 2025, either gets reaffirmed, raised, or becomes the source of the next leg down if any single hyperscaler blinks. Google Cloud's 82% growth sets a high bar for Azure and AWS, but it also validates the demand narrative. We do not expect any hyperscaler to cut guidance, the supply constraint is the binding variable, not demand.
3. NVDA as the Capex Proxy. NVIDIA does not report next week, but its stock price is effectively a real-time poll on whether the hyperscaler capex guidance is being believed. The chip complex was down hard Thursday before partially recovering Friday. The S&P 500 is now 2.6% below its June 2 record close after two consecutive weekly declines, with the index having slipped below its 50-day moving average. A confirmed hold above the 200-day for NVDA and AVGO next week, watched closely, determines whether the current repricing is a dip inside the trend or the beginning of a more extended digestion phase.
Forward Thesis by Pillar: Conviction and What Would Change It
AI Chips / Hardware (NVDA, AMD, AVGO): Conviction remains high. This week's INTC blowout and the AMD-Microsoft partnership expansion are additive evidence that AI chip demand is broadening: GPUs first, then custom silicon, now CPUs. Intel's CFO noted that CPUs have reached nearly 1:1 parity with GPUs in certain agentic AI workloads, with CPUs increasingly required to facilitate data movement, memory retrieval, and agentic workflows. That is the structural underpinning of the multi-vendor thesis. What would change it: a hyperscaler capex guidance cut next week, or NVDA breaking below its 200-day on confirmed closes. Neither has happened. The sell-the-news dynamic in INTC and the broader chip complex this week is a price disconnect, not a thesis break.
AI Software / Platforms (PLTR, META): Google Cloud +82% is the single most important data point for this pillar this quarter. It confirms that enterprise AI workloads are converting from pilots to production at accelerating rates, the exact demand substrate that META's AI ad monetization platform and PLTR's enterprise AI OS are built on. META reports next week. Meta raised its full-year 2026 capex guidance to $125 to $145 billion, citing higher component pricing and additional data center costs. If the revenue growth that Alphabet delivered on the cloud side translates to Meta's AI-native ad platform, the stock's recent pullback will look like the last clean entry of the cycle. What would change it: a deceleration in Meta's AI-driven ad revenue growth, or forward guidance implying the capex is running ahead of monetization capacity.
Nuclear / Power Infrastructure (VST, CEG, LEU): Conviction elevated. Oil at $100 is the unambiguous structural accelerant for zero-fuel-cost nuclear baseload. The expectation is that the Fed keeps rates on hold for the rest of 2026. A hold-rate environment combined with contracted power purchase agreements and Hormuz-driven fossil fuel risk premium is exactly the scenario where VST and CEG's long-duration contracted cash flows command a premium multiple. What would change it: a sustained ceasefire in the Strait of Hormuz driving Brent back below $75, which would reduce the urgency premium on domestic power infrastructure, though the AI data center demand driver remains intact regardless of oil price.
Digital Finance Rails: Thesis intact, catalyst timeline delayed. The missed July 18 deadline on GENIUS Act rulemaking pushes the near-term catalyst window toward Q4. Monitoring the pace of rule finalization. What would change it: explicit legislative reversal of the GENIUS Act framework (not on the table), or a sustained adverse ruling on stablecoin issuer reserve requirements that undermines the institutional adoption case.
Specialized GPU Cloud (CRWV): The most fragile name in the complex by price action. Tracking closely. The NVIDIA-backstopped contracted infrastructure thesis is intact, the hyperscaler capex prints next week are the direct read-through. A strong combined print from Meta, Microsoft, Amazon, and Apple on AI infrastructure demand is the catalyst that stabilizes the contracted GPU cloud names. Watching for confirmation.
The fund's singular read entering next week: the earnings data is confirming the AI revenue thesis at the infrastructure layer; the market is repricing multiples in response to capex escalation and geopolitical oil risk. Those are two different things. Google Cloud at 82% growth does not coexist with a broken AI investment thesis. It coexists with a market that has temporarily lost patience for the gap between spend and return, a gap that Google Cloud just proved is closing faster than the bears projected. The hyperscaler grand finale next week will either confirm that read or force a reassessment. We know which way the data is pointing.
This document is for informational purposes only and does not constitute investment advice or a solicitation to buy or sell any security. The Nova X Convergence Fund letters reflect the analytical views and investment thesis of the fund's research team and should not be relied upon as the basis for any investment decision. Past performance is not indicative of future results. All investments involve risk, including the potential loss of principal. The information contained herein is believed to be from reliable sources but is not guaranteed as to accuracy or completeness. Published July 25, 2026.