Key Takeaways
- Nvidia shares advanced 15% in the last 30 days, though AMD’s 38% and Intel’s 56% gains outpaced the chipmaker during this period.
- Rick Schafer from Oppenheimer maintains an Outperform rating on Nvidia with a $265 price objective, naming it his preferred semiconductor investment.
- The company’s Blackwell Ultra (GB300) NVL rack systems reportedly lead rival AI chip offerings by two full generations.
- NVDA shares currently trade at 17x projected 2027 earnings, representing a discount to the semiconductor sector’s 20x average multiple.
- The company’s revenue trajectory climbed from $27 billion in fiscal 2023 to approximately $216 billion in fiscal 2026, with analyst projections reaching $480 billion by fiscal 2028.
Nvidia has posted solid gains among chip manufacturers during the recent month-long period — though not the strongest. NVDA registered a 15% increase while AMD rocketed 38% higher and Intel soared 56%. This performance divergence has sparked investor discussion, yet one Wall Street analyst suggests the gap doesn’t warrant concern.
Rick Schafer at Oppenheimer assigns Nvidia an Outperform rating alongside a $265 price objective. In contrast, he rates both AMD and Intel as Perform. Schafer identifies Nvidia as his leading semiconductor recommendation as earnings announcements approach.
The stronger recent performance from AMD and Intel stems from heightened attention on CPUs designed for AI server configurations — a segment distinct from Nvidia’s GPU-centric operations. Intel additionally benefited from a favorable mention in Barron’s recent stock selections.
Meanwhile, Nvidia’s narrative continues to develop. NVDA changed hands near $198.60 during Friday morning pre-market activity on April 17, showing a 0.2% uptick.
Trading Multiple Sits Below Industry Norm
Schafer’s analysis highlighted Nvidia’s Blackwell Ultra (GB300) NVL rack technology as maintaining a two-generation advantage over competing solutions. He further noted that Nvidia presently commands approximately 17x his earnings projection for 2027 — underneath the sector’s 20x average — representing an appealing entry point for a firm with such market leadership.
The shares have climbed roughly 75% across the trailing twelve months. With a backward-looking P/E near 41x, some market participants have expressed skepticism. However, Trefis analysts contend this valuation remains justified considering the expansion potential ahead.
Revenue figures illustrate this trajectory. Nvidia expanded from $27 billion in fiscal 2023 to nearly $216 billion in fiscal 2026 — approximately 8x expansion. Consensus estimates now project $480 billion in top-line revenue by fiscal 2028.
Two primary elements support this forecast. First is the transition from AI model training to inference workloads. Training occurs intermittently; inference runs continuously. As agentic AI applications proliferate, computational requirements intensify. Organizations already integrated into Nvidia’s CUDA platform encounter substantial switching barriers, creating customer retention.
National AI Infrastructure and Profitability Dynamics
The second expansion catalyst involves Sovereign AI initiatives. Nations globally are establishing domestic AI computing infrastructure, and Nvidia’s CUDA-centered technology stack positions it favorably. During fiscal 2026, Nvidia’s sovereign AI segment tripled, exceeding $30 billion.
Regarding profitability, Nvidia posted 54% net margins in fiscal 2026, advancing from 31% in fiscal 2023. AMD, for comparison, operates around 20% margins. Trefis forecasts margins sustaining approximately 52% despite evolving product composition.
Should revenue achieve $575 billion with margins maintaining the 52% level, this suggests net earnings near $300 billion — roughly 2.5 times the $117 billion delivered in fiscal 2026.
Applying a trailing P/E ratio of 25x to that income figure, Trefis calculates a potential market capitalization of $7.5 trillion, which could drive the stock price toward $300.
For perspective, Cisco currently trades near 22x trailing earnings. Microsoft exceeds 27x.
Nvidia’s forthcoming architecture, Vera Rubin, will succeed Blackwell and aims to enhance inference performance while reducing per-token computational costs.
