Goldman Sachs Reiterates Buy on NVIDIA, Bullish on Data Center Revenue Beat

Overview

This report presents Goldman Sachs’ forward-looking analysis of NVIDIA’s (NVDA) Q1 FY2026 earnings. The report argues that, despite strong supply chain execution and increased capital expenditures by major cloud providers driving elevated market expectations, NVIDIA remains well-positioned to deliver a 'beat-and-raise' quarter. Goldman Sachs raises its average 2026/2027 EPS forecasts by ~12%, substantially above consensus, and reiterates its Buy rating and $250 target price. Key focal points include whether data center revenue can exceed the $1 trillion cumulative guidance underpinned by Blackwell and Rubin platforms, and the incremental uplift from Agentic AI on CPU-related business.

Core views

Earnings Expectations and Valuation Re-rating: Goldman Sachs expects NVIDIA’s Q1 revenue to exceed consensus by ~$2 billion and Q2 revenue to be ~3% above consensus; EPS for Q1 and Q2 is projected to be 7% and 8% above consensus, respectively. Based on management’s positive commentary regarding the visibility into $1 trillion in data center revenue from Blackwell and Rubin platforms, Goldman Sachs raises its 2026 and 2027 EPS forecasts to levels 14% and 34% above consensus, respectively. Although NVIDIA’s stock has recently underperformed peers and trades at a historical valuation discount, the report suggests that valuation multiples could expand if investors observe evidence of improved profitability among hyperscalers supporting sustained capex, broadening enterprise adoption signaled by Agentic AI proliferation, and heightened visibility into deployments by non-traditional customers. Core Growth Drivers: Data Center and Agentic AI: Investors will closely monitor upside potential to NVIDIA’s $1 trillion cumulative data center revenue guidance for 2025–2027—announced at GTC 2026—for Blackwell, Blackwell Ultra, and Rubin products. Potential upside sources include Rubin Ultra in 2027, Vera pure-CPU racks starting in 2026 and beyond, and inference-optimized configurations (e.g., Rubin-CPX and Groq 3 LPX). Additionally, Agentic AI is expected to strongly boost server CPU demand, with NVIDIA’s pure-CPU racks slated to begin shipping in H2 2026—creating a new growth vector for accelerator market share and total addressable opportunity over the medium term. Competitive Landscape and Margin Outlook: Facing increasing reliance by hyperscalers on custom XPUs, Goldman Sachs expects management to update its views on competition from custom chips. NVIDIA is expected to emphasize its leadership in delivering the lowest inference cost via annual product iteration, with data showing ~10x cost improvement per generation for Blackwell. On margins, despite rising component costs and accelerated Rubin platform rollout in H2, Goldman Sachs expects NVIDIA to reaffirm its full-year FY2026 gross margin guidance in the mid-70% range.

Analysis framework

Goldman Sachs’ analytical framework centers on 'expectation gaps' and 'catalysts.' First, it quantifies near-term earnings upside (e.g., $2bn revenue beat, 7–8% EPS beat) by comparing internal forecasts against Street consensus, establishing short-term earnings support. Second, it deconstructs long-term growth logic—broadening the lens beyond GPU compute alone to assess incremental CPU demand driven by Agentic AI and demand trends from non-traditional customers (e.g., sovereign AI, OpenAI)—to validate the breadth of growth. Finally, it evaluates earnings quality and valuation re-rating potential amid high base effects by assessing competitive dynamics (custom ASIC vs. GPU) and cost structure (gross margin resilience).

Methodology notes

• Valuation MethodologyP/E and PEG Valuation

  Price-to-earnings multiple applied to normalized EPS

  The report applies a 30x P/E multiple to Goldman Sachs’ normalized EPS forecast of $8.25 for FY2026 to derive the target price. This methodology is appropriate for high-growth technology stocks, assigning a premium multiple to reflect future earnings growth potential.

• Industry/Vertical Analysis FrameworkSupply-demand framework

  Assessing industry health through supply-side signals (e.g., TSMC, SK Hynix earnings) and demand-side indicators (e.g., cloud provider capex)

  The report uses strong upstream supplier performance (e.g., TSMC, SK Hynix) as a supply-side signal, combined with U.S. hyperscaler capex increases for FY2026, to jointly assess semiconductor industry tightness and NVIDIA’s shipment outlook.

• Event-Driven and Behavioral FinanceExpectation Gap / Expectation Management

  Focus on key non-financial guidance updates during earnings calls that may shift market expectations

  The report explicitly highlights investor attention to non-financial metrics during the earnings call—such as upside to the $1 trillion data center revenue guidance and updated commentary on Agentic AI’s impact on CPU demand—as critical catalysts driving near-term price volatility.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• NVIDIA CORP (NVDA.US)

  Direct beneficiary as the core AI compute supplier, benefiting from increased data center capex and new demand driven by Agentic AI.

  Strengths

  Leading Blackwell/Rubin platform products, significant generational cost improvements (~10x), robust software ecosystem and strong customer stickiness.

  Weaknesses

  Trading at a relatively low historical valuation, facing intensifying competition from custom ASICs and rising component cost pressures.

  Comparison

  Has underperformed peers recently but exhibits significantly higher earnings forecasts than consensus, offering valuation re-rating potential.

  Risks

  Slowing AI infrastructure spending, market share loss or margin erosion due to competition, supply chain constraints.

Key data

• Target Price$250Based on 30x P/E and $8.25 normalized EPS forecast
• Current Stock Price$207.83As of market close on May 6, 2026
• Implied Upside20.3%Calculated from target price relative to current stock price
• Q1 Revenue Expectation Gap+$2BGoldman Sachs expects ~$2 billion above consensus
• Q1 EPS Expectation Gap+7%Goldman Sachs expects 7% above consensus
• 2026/2027 EPS Forecast RevisionAvg. +12%Goldman Sachs raised forecasts, now 14% and 34% above consensus, respectively
• FY2026 Gross Margin GuidanceMid-70% rangeExpected to be reaffirmed despite rising cost pressures
• Blackwell-Generation Inference Cost Improvement~10xSignificant reduction in inference cost per generation

Impact & implications

The report contends that if NVIDIA can demonstrate that hyperscaler profitability improvements sustain capex and that Agentic AI achieves broad enterprise adoption, its valuation multiple could expand from its current historically discounted level. For investors, the focus should shift from GPU unit shipments alone toward broader ecosystem monetization—including CPU business growth and penetration among non-traditional customers (e.g., sovereign AI). Sustained demonstration of NVIDIA’s cost leadership in custom chip competition—such as the ~10x cost improvement enabled by Blackwell—would reinforce its long-term moat.

Risks

• Slowing AI infrastructure spending
• Market share erosion due to intensified competition
• Margin erosion due to intensified competition
• Supply chain constraints

What to watch

• Monetization trends among hyperscalers and adoption velocity of Agentic AI across consumer and enterprise workloads
• Execution by emerging neo-clouds and non-hyperscale customers
• Competitive intensity from custom XPUs and AMD solutions
• Impact of rising memory and other input costs on gross margin