The earnings calendar has been building toward this afternoon for three weeks. Alphabet reports after the close at approximately 1:00 PM PT, Tesla follows. Both report into a market that just staged its biggest two-day semiconductor recovery in months, with MU up $155 from its recent lows, the SOX off its bear market floor, and institutional capital visibly returning to AI hardware names. Whether that recovery sustains or collapses depends almost entirely on what these two companies say this afternoon. Alphabet in particular is what Nvidia is to the supply side of AI: the single clearest read on whether the demand that justifies the entire infrastructure buildout is real, growing, and monetizable.
What Alphabet Actually Needs to Deliver
The consensus numbers are straightforward. Total revenue of approximately $117.1 billion represents 21% year-over-year growth. EPS consensus sits at $2.88. Advertising revenue is expected to rise 14% to $81.3 billion. Google Cloud is projected at approximately $22.4 to $22.8 billion, representing roughly 60 to 67% year-over-year growth from $20.03 billion in Q1.
The headline revenue beat is likely. Alphabet has beaten consensus for eight consecutive quarters, and Q1's Q1 beat of $109.9 billion against $107.2 billion expected set a constructive precedent. What the market is not pricing on the headline, and what will actually determine whether the stock moves up or down this afternoon, is the resolution of three specific tensions.
The first is the capex question. Alphabet raised its 2026 capital expenditure guidance from $175 to $185 billion in Q1 to $180 to $190 billion at the April earnings call, and Q1 alone consumed $35.7 billion. FactSet consensus now expects $187.1 billion for the full year. Management indicated on the Q1 call that 2027 capex would "significantly increase" beyond 2026. Some analysts expect another upward revision this afternoon toward $190 to $200 billion. The market's response to that number depends entirely on what accompanies it. Capex rising to $190 billion alongside Google Cloud growth of 65% and margin expansion is a demand confirmation that justifies the spending. Capex rising to $190 billion alongside any sign of cloud deceleration or margin compression is the overbuild signal that has punished hyperscalers in prior cycles.
The second is Google Cloud's backlog conversion. The cloud backlog nearly doubled to $462 billion. What matters is the pace at which that backlog converts to recognized revenue. If Q2 Cloud revenue comes in at or above $22.5 billion and management signals that backlog conversion is accelerating, the spending math validates. If Cloud disappoints relative to the backlog size, the market draws the inference that customers are committing capacity but slowing deployment.
The third is search resilience against AI cannibalization. AI Overview, Alphabet's AI-powered search experience, is now deployed across more than a billion users. The structural fear is that AI summaries replace the click-through behavior that generates search advertising revenue. Q2 advertising revenue at $81.3 billion consensus represents a meaningful test of whether that cannibalization is materializing at scale.
The Gemini 3.5 Pro Delay and the New AI Chip
Two specific developments heading into this afternoon add complexity to the narrative. Alphabet reportedly delayed the launch timeline for Gemini 3.5 Pro, its flagship AI model, which has already rattled the stock in recent weeks. The delay matters not just as a product timeline issue but as a signal about Alphabet's competitive positioning relative to Anthropic, OpenAI, and the Chinese open-source models that produced the Kimi K3 moment last Friday. If management addresses the delay on the earnings call with a credible updated timeline and evidence of interim model performance, the market will move on. If the call produces vague reassurances without specifics, the concern deepens.
Separately, Alphabet is reportedly planning a new custom AI chip. Broadcom co-designs Alphabet's existing TPUs and represents the single largest revenue concentration in Broadcom's AI custom silicon business. Any disclosure about next-generation chip architecture this afternoon would directly affect AVGO's forward revenue visibility and re-price the entire custom silicon supply chain.
Who Benefits From Alphabet's Capex
If Alphabet raises its capex guidance this afternoon and provides constructive cloud and search commentary, the direct beneficiaries extend well beyond the stock itself. Broadcom captures the custom silicon revenue from TPU co-design. Arista Networks provides the networking infrastructure for Google's data center clusters. Vertiv supplies power and cooling systems. The semiconductor equipment names, ASML, Applied Materials, KLAC, and Lam Research, benefit from the fab expansion that feeds into Alphabet's server build. Nvidia GPU supply runs through Alphabet's infrastructure alongside its own TPUs. The entire AI infrastructure supply chain gets a demand signal from Alphabet that it cannot get from any other single source.
This is the read-through logic that makes this afternoon more important than a single company's quarterly results. Alphabet's capex number is the most direct available confirmation of whether the $725 billion-plus collective hyperscaler AI spending commitment that Micron and SK Hynix are building their contracted revenue models around is intact, accelerating, or moderating.
Tesla: The Physical AI Companion Test
Tesla reports simultaneously, with an investor call at 2:30 PM PT versus Alphabet's 1:30 PM PT. Consensus expects $26.4 billion in revenue, up roughly 17% year over year, and adjusted EPS of $0.53. Q2 deliveries of 480,126 units came in above prior market expectations, which means the revenue line is unlikely to be the problem.
The variables that determine the stock's reaction are automotive gross margin excluding regulatory credits, free cash flow, and the specificity of Musk's commentary on Robotaxi deployment and Optimus robotics commercialization. Market consensus expects automotive gross margin excluding credits of approximately 18.1%. Two consecutive quarters of margin compression have been the primary financial concern, and a recovery toward or above 18% would remove the most obvious objection to the stock's AI narrative premium.
Robotaxi matters because it is the monetization thesis that makes Tesla a technology company rather than a car company at its current valuation. If Musk provides a credible deployment timeline with specific cities, revenue models, and regulatory progress, the stock's AI infrastructure premium survives. If the commentary remains high-level without operational specifics, the premium faces continued compression.
Texas Instruments also reports this afternoon, providing the analog semiconductor read-through. TI's industrial and automotive chip demand signals are the most direct available indicator of whether the non-AI semiconductor cycle has genuinely recovered from its 2024 to 2025 trough. A strong TI print would confirm that memory and AI hardware strength is accompanied by genuine broad industrial recovery rather than an isolated AI-driven spike.
The Broader Stakes for the AI Trade
The sequence of this afternoon's reports carries a specific market logic. Alphabet first, because cloud growth and capex guidance set the AI infrastructure demand picture. Tesla second, because Robotaxi and Optimus set the physical AI story. TI simultaneously, because industrial chip demand sets the macro context.
If Alphabet delivers Cloud revenue above $22.5 billion, maintains or expands operating margins, and raises capex guidance alongside constructive demand commentary, the market will treat the past three weeks of semiconductor selling as a sentiment-driven overcorrection against intact fundamentals. Micron's $155 recovery from its lows, BofA's US 1 List addition, Morgan Stanley's DRAM price upgrade, and the UBS momentum unwind call all converge on this moment. This afternoon either validates the institutional thesis that bought the dip, or exposes it as premature.
The alternative scenario, where Alphabet's cloud growth disappoints or capex rises without a proportional demand confirmation, extends the uncertainty about whether the AI infrastructure buildout is generating returns at the pace the spending requires. That scenario does not end the AI trade. It extends the period of elevated volatility while the market waits for more evidence.
Alphabet decides this afternoon whether AI can still be traded. Tesla decides whether physical AI can still be believed. The answer to both arrives before the close of the US trading day.
Explore more exclusive insights at nextfin.ai.
