Nvidia just had another blowout quarter, leaving analysts even more bullish on the chipmaker

Nvidia posted fiscal Q2 revenue of $96.22B, up 106% YoY and above the $92.17B LSEG consensus, alongside adjusted EPS of $2.22 topping estimates. The company also guided to ~70% fiscal 2028 revenue growth as AI hits an “inflection point,” with analysts pointing to demand remaining “off the charts” and management addressing gross margin concerns (guiding GM to ~72–73% vs ~75% amid higher memory costs). Off-hours, the stock was up about 7% and is up ~12% YTD, with multiple Street firms raising/affirming targets (e.g., Bernstein PT to $400 from $315, implying ~91% upside).

Analysis

The key market read-through is that AI demand is no longer just a semiconductor story; it is migrating into a broader capacity bottleneck across memory, advanced packaging, power, and networking. That favors the most constrained suppliers first, not necessarily the most exposed ones, so the incremental winners are TSM, MU, ANET, and power/cooling beneficiaries rather than a blanket move across semis. The corollary is that lower-quality competitors and custom-silicon efforts face a harsher bar: if the platform leader is still taking share while demand outruns supply, alternative architectures need a much clearer cost/performance edge to matter.

Near term, the trade is momentum plus estimate revision risk, but the clock matters. Over days to weeks, a gap-up can squeeze short interest and force buy-side model resets higher; over 1-3 months, the real catalyst is whether hyperscaler capex guides re-accelerate enough to validate the new forward numbers. The main reversal trigger is not a bad quarter from NVDA, but evidence that demand is being pulled forward by vendor financing and that end customers are stretching balance sheets rather than expanding true deployment capacity.

The consensus may be overestimating how durable the margin story is and underestimating the credit risk embedded in the ecosystem. If growth is being subsidized through infrastructure financing, the next stress point is quality of demand: any slowdown in startup funding, a pause in hyperscaler capex, or a slip in Rubin timing would compress the multiple quickly. Conversely, if NVDA can keep guiding up while gross margin only modestly compresses, the stock still has room to outperform, but the cleaner expression is likely relative value rather than an outright chase after a big post-earnings move.

Key Decisions for Investors

• Long NVDA / short AMD for 1-3 months: NVDA has the clearest estimate-revision path and platform lock-in; AMD is the most obvious substitute if buyers rotate on valuation. Falsify if AMD wins a major hyperscaler design cycle or NVDA fails to hold the post-earnings gap over the next 2 sessions.
• Buy SMH or TSM on weakness over the next 1-4 weeks: the real scarcity value sits in packaging and manufacturing capacity, not just in the GPU itself. Exit if TSM commentary shows packaging utilization normalizing faster than expected or if AI capex growth slows in upcoming hyperscaler updates.
• Add MU as a 3-6 month expression of the HBM pricing squeeze: memory inflation is a headwind for NVDA margins, but it is a larger tailwind for memory vendors with allocation leverage. Cut if HBM ASPs flatten or if NVDA gross margin guidance drops below the low-70s again.
• Use NVDA call spreads rather than outright stock for the next 4-8 weeks: implied volatility is likely rich after the print, and the stock needs follow-through from analyst estimate upgrades to sustain upside. If the shares stall after the opening gap, take profits quickly and avoid paying up for decay.
• Watch the AI financing channel as a risk alert, not a buy signal: if vendor funding becomes a larger part of demand, the 6-18 month risk shifts to counterparty quality and bad-debt surprises across the ecosystem. That would be a cue to trim the more speculative AI infra names first.