Wolfe Research reiterates Nvidia stock rating on datacenter upside By Investing.com

$1 trillion: Nvidia management and analysts are citing visibility into roughly $1 trillion of Blackwell, Blackwell Ultra and Rubin revenue through calendar year 2027. Wolfe Research reiterated an Outperform with a $275 price target and projects base/bull EPS of ~$12.50/$14 (versus consensus $8.23 for fiscal 2027), implying ~14–17% upside to datacenter estimates and potentially >25% upside in a bull case; other analyst moves include Raymond James raising its PT to $323, Truist to $287, and Argus at $220. The commentary, combined with ~65% trailing 12-month revenue growth and continued product ramps (Rubin Ultra, STX, RTX, LPX), reinforces bullish positioning on Nvidia and the AI/data-center market.

Analysis

NVIDIA’s cadence of new architectures is reshaping buyer behavior: hyperscalers and system OEMs increasingly plan multi-year refresh cycles around accelerator roadmaps rather than CPU replacement cadence. That reallocates a larger share of near-term datacenter wallet to GPU-led stacks, creating an outsized pull-through for advanced packaging, HBM memory, and test/equipment suppliers while compressing incremental CPU-only spend.

Second-order supply-chain winners are equipment and specialty memory vendors that face a multi-quarter lead time to scale; constrained supply in those nodes will sustain vendor pricing power into the next procurement wave. Conversely, firms that rely on legacy CPU ASP expansion or that sell commoditized server components without accelerator integration risk margin pressure as customers trade up to integrated system buys.

Key risks span three horizons: days (market reactions to quarterly guide and product cadence commentary), months (inventory digestion and OEM channel fills), and years (software ecosystem lock-in and potential competitive architectural breakthroughs). Geopolitical export policy or a meaningful yield setback on advanced packaging would be the fastest paths to a material reset of expectations; conversely, sustained customer integration wins would propagate upside to suppliers beyond the semiconductor die makers themselves.

Key Decisions for Investors

• Buy NVDA Jan-2028 call spread (LEAP): buy 1x Jan-2028 1) near-the-money call and sell a higher strike to finance ~50-70% of premium. Size at 3-5% of tech sleeve. Rationale: multi-year product-led TAM expansion with capped premium; downside = full debit, upside = substantial asymmetric payoff if multi-year adoption holds.
• Buy Lam Research (LRCX) or Applied Materials (AMAT) 9–12 month exposure: accumulate on any post-earnings pullback and hedge with a 6–9 month 10–15% OTM protective put. Rationale: equipment demand is the choke point for ramp — outsized upside if capex stays elevated; hedge limits drawdown if end-market demand softens.
• Pair trade — long NVDA equity (or LEAP) / short INTC equal notional for 12 months: size modestly (2–4% net notional). Rationale: asymmetric market share transfer toward accelerators should favor NVDA over legacy CPU incumbents; risk if Intel executes new integrated accelerator wins or if NVDA guidance disappoints.
• Tactical long on systems/OEMs with accelerator integration (e.g., HPE/DELL) for 6–12 months: buy on weakness around macro noise. Rationale: systems vendors re-rating as attach rates and ASPs per rack increase; downside limited by diversified services revenue but monitor customer inventory builds closely.