Nvidia sales jump 85% as AI boom fuels record growth

Nvidia posted a very strong quarter, with revenue up 85% year over year to $81.6 billion and adjusted operating profit up nearly 150% to $53.5 billion, while net income rose to $58.3 billion. The company guided current-quarter revenue to about $91 billion, above analyst estimates, and said demand remains strong despite ongoing supply constraints for next-generation Vera Rubin chips. Investors also remain focused on China risk, where a 25% fee on some chip sales and uncertain approval from Beijing could affect future growth.

Analysis

The market is still treating Nvidia like a single-name earnings story, but the more important signal is that it remains the bottleneck supplier to an entire capital cycle. That means the real beneficiary set is broader than semis: foundry capacity, advanced packaging, HBM memory, substrate vendors, and power/cooling infrastructure should keep seeing orders ratchet higher for at least the next 2-3 quarters as customers pre-buy against the next platform transition.

The key second-order risk is not demand decay; it is margin and mix compression from supply normalization and policy friction. If the next-generation ramp stays constrained while customer budgets shift from emergency buildout to optimization, the market may start discounting a slower growth cadence even if absolute revenue remains huge. China is the cleanest near-term overhang: any incremental access is likely to be low-quality, fee-laden volume with lower profitability and higher headline risk, so the stock can underperform on “good” news if investors conclude the region is structurally less monetizable.

Consensus is likely underestimating how much of AI spend is still being funded out of hyperscaler capex, not end-demand monetization. That makes the setup vulnerable to a 6-12 month digestion phase if cloud providers pause to integrate what they already bought. On the upside, the CPU and robotics adjacency matters less for immediate earnings but matters a lot for narrative durability: it broadens Nvidia from an AI accelerator supplier into a compute platform, which justifies a higher terminal multiple if execution holds.

The contrarian takeaway is that the stock’s short-term reaction is more about positioning than fundamentals; when expectations are this elevated, even exceptional numbers can produce a flat-to-down tape. That creates opportunity to own the ecosystem rather than chase the flagship name, because the supply chain tends to lag the headline product cycle by 1-2 quarters and has less policy headline risk.

Key Decisions for Investors

• Prefer long positions in the NVDA supply chain over chasing NVDA here: long TSM, ASML, and a basket of HBM/packaging beneficiaries on a 3-6 month horizon; risk/reward is better because the market still underprices the duration of AI capex while the headline name is already priced for perfection.
• Buy NVDA downside protection into the next 4-8 weeks via put spreads rather than outright shorts; the stock is vulnerable to any guidance normalization or China headline, but the secular trend remains intact, so defined-risk hedges are superior.
• Pair trade: long semiconductor equipment / infrastructure basket vs. short a broad tech ETF over 3-6 months; if AI spending decelerates at the margin, the first-order hit is to sentiment, but the second-order hit is to spending efficiency, which should favor picks-and-shovels over high-multiple software.
• If NVDA sells off on the print but holds above its 50-day moving average after 1-2 sessions, fade the weakness with a tactical long; that setup would imply a positioning flush rather than a fundamental break, offering a favorable 2:1 upside/downside for a 2-4 week trade.
• Do not add aggressively on China easing headlines alone; any incremental revenue from that channel should be treated as lower-quality and policy-dependent, so only size longs if the market confirms a broader reacceleration in order backlog.