Nvidia (NVDA) Q1 2027 Earnings Transcript

NVIDIA reported record Q1 revenue of $82 billion, up 85% year over year and 20% sequentially, with data center revenue of $75 billion (+92% YoY) and free cash flow of $49 billion. Management raised the quarterly dividend to $0.25 per share, authorized an additional $80 billion buyback, and guided Q2 revenue to $91 billion (+/-2%), while explicitly excluding China data center compute revenue from outlook. The call highlighted accelerating Blackwell demand, Vera/VeraRubin product ramps, and expanding AI infrastructure adoption across hyperscale, AI-native, sovereign, and edge markets.

Analysis

The market is still underpricing how much of NVDA’s growth is shifting from a single hyperscaler CapEx cycle into a broader, more resilient infrastructure buildout. The segmentation change is not cosmetic: it signals that the next leg is less about a handful of cloud spenders and more about fragmented, regulated, and latency-sensitive deployments where NVDA’s full-stack moat is strongest and competition is structurally weaker. That should extend durability of revenue growth even if hyperscaler budgets normalize, because ACIE and edge create a longer tail of smaller but stickier order flow.

The bigger second-order effect is that NVDA is increasingly monetizing the entire AI value chain, not just accelerators. Vera turns the CPU from a commodity into an orchestration layer for agentic workloads, which expands attach rates across networking, security, storage, and software tooling. That is bullish for adjacent beneficiaries like CRWV and SNPS in the near term, but it is also a warning for MSFT/AMZN/GOOGL/META: their AI CapEx intensity likely stays elevated because NVDA is pulling more of the economics into the stack while making it harder to substitute with in-house silicon.

The main risk is not demand; it is digestion. Supply commitments, record OpEx growth, and extreme customer concentration mean the stock is vulnerable if execution slips on VeraRubin ramp timing or if token economics compress faster than expected. China remains a free call that is not in the guide, so any policy thaw becomes upside optionality, but the base case should assume no help there for several quarters. In the meantime, the pricing of H100/A100 rentals suggests a still-hot secondary market, which usually precedes another leg of hardware demand rather than a peak.

Key Decisions for Investors

• Stay long NVDA into the next 1-2 quarters, but fund it by trimming some high-beta semis; use dips post-rally to add, since the broader ACIE/edge mix lowers dependence on hyperscaler CapEx alone.
• Pair trade: long NVDA / short a basket of in-house AI silicon proxies (particularly names where custom chip narratives imply substitution risk) for 3-6 months; the catalyst is Vera/VeraRubin proving that NVDA can grow attach, not just unit share.
• Add CRWV on pullbacks over the next 1-2 months as a second-order winner from NVDA’s infrastructure expansion; high-risk/high-reward given customer concentration, but it should benefit from the same capacity buildout.
• Buy MSFT/AMZN/GOOGL on weakness only if cloud spend selloff follows NVDA strength; their AI spend is likely to remain elevated, but expect margin pressure over the next 2-4 quarters as they compete for scarce compute.
• Use a medium-dated NVDA call spread rather than outright stock for event capture into the VeraRubin ramp window (Q3-Q1 next year), since execution risk exists but the upside from another demand re-acceleration remains meaningful.