NVIDIA’s expanded supply agreement with AWS demonstrates that enterprise and cloud provider AI infrastructure procurement remains robust well beyond initial deployment phases.

Model NVIDIA’s growth assumptions yourself with TIKR’s Guided Valuation Model (It’s free) >>>

NVIDIA (NVDA) reported one of the largest quarters in corporate history, underscoring how high the market’s expectations have risen. Fiscal Q2 revenue reached $96.2 billion, a 106% year-over-year increase, with data center revenue alone hitting $89 billion, up 117%. Adjusted earnings per share of $2.22 also exceeded analyst estimates.

NVIDIA manufactures the graphics processing units (GPUs) that power most of the world’s artificial intelligence systems. These chips train and operate the large language models behind applications like ChatGPT, and demand has surged as enterprises accelerate AI infrastructure buildouts. Given NVIDIA’s dominant market position, its financial results are widely regarded as a bellwether for the broader AI sector.

For Q3, NVIDIA guided revenue to $108 billion, plus or minus 2%, a range comfortably above pre-report analyst expectations. Chief Financial Officer Colette Kress noted that demand continues to outpace supply, with the company actively working to bridge the gap through fiscal 2028.

CEO Jensen Huang summarized the current landscape succinctly. “AI has reached its inflection point,” Huang stated during the earnings call, characterizing the shift as computing power finally translating into direct customer revenue rather than speculative investment.

NVIDIA disclosed a $0.4 billion charge related to excess inventory, specifically tied to older H200 chips following reduced Chinese demand under U.S. export restrictions. Hopper chip shipments to China accounted for less than 1% of data center revenue this quarter. Management has excluded any China compute revenue from its forward guidance. Should supply constraints ease as anticipated, the current growth trajectory could sustain well into next year.

Compare your own AI demand assumptions against NVIDIA’s guidance (It’s free) >>>

Under the valuation model’s assumptions, realized through January 31, 2029, the stock is modeled using:

Based on these inputs, the model projects a target price of $412.46, representing 89.6% total upside from the current share price and a 30.1% annualized return over a 2.4-year horizon.

This places NVIDIA firmly in undervalued territory according to the model’s thresholds. Notably, the exit multiple assumption is set at just 18x, significantly below NVIDIA’s current forward P/E, which ranges from the mid-teens to low twenties depending on the estimate. Even a conservative multiple yields a substantial target price, driven primarily by the steep revenue and margin growth embedded in the model.

The bull case hinges on NVIDIA’s expanding revenue opportunity per gigawatt of data center capacity. Management reports this metric has risen from approximately $18 billion during the Hopper generation to $25 billion with Blackwell, and $40 billion with the forthcoming Vera Rubin platform. Rather than declining as typically seen in mature technology cycles, each successive architecture captures greater spending per unit of infrastructure.

Conversely, the bear case focuses on geopolitical exposure and customer concentration. Approximately half of NVIDIA’s data center revenue derives from a narrow cohort of hyperscale cloud providers; a reduction in their capital expenditures would disproportionately impact NVIDIA compared to more diversified peers. The H200 inventory writedown underscores how geopolitical shifts can outpace product development cycles.

Currently, the growth narrative outweighs the risks outlined in the model’s projections. However, investors should closely monitor capital expenditure commentary from NVIDIA’s largest clients in upcoming quarters.

Estimate a company’s fair value instantly (Free with TIKR) >>>

Advanced Micro Devices (AMD), NVIDIA’s closest direct competitor in AI accelerators, saw data center revenue more than double to $6.7 billion last quarter, up from $3.2 billion a year prior. While impressive, AMD’s data center segment remains a fraction of NVIDIA’s $89 billion total. AMD currently trades at a steep forward P/E of approximately 83x, pricing in investor optimism that its EPYC processors and Instinct GPUs will gradually close the performance gap.

Broadcom (AVGO) pursues a different strategy, providing custom AI chips and networking equipment rather than standardized GPUs. Broadcom’s AI semiconductor revenue surged 143% year over year to $10.8 billion last quarter, while total company revenue increased 48% to $22.2 billion. Trading at roughly 37x forward earnings—a discount to AMD but still a premium valuation—the stock reflects growing custom silicon partnerships with major cloud providers.

NVIDIA maintains a decisive scale advantage, with its data center business roughly eight times larger than Broadcom’s total AI revenue. The central competitive dynamic is not whether AMD or Broadcom can surpass NVIDIA outright, but whether they can capture sufficient share in networking and custom silicon to moderate NVIDIA’s growth trajectory over the coming years.

Track Q2 earnings in late August for data center revenue trajectory, gross margin sustainability, and any update on next-gen Blackwell ramp >>>

The AWS partnership represents the most tangible near-term catalyst. NVIDIA and Amazon Web Services announced plans to deploy two million additional GPUs across AWS infrastructure in 2027 and 2028, building on over one million GPUs already scheduled for deployment beginning in 2026. This agreement alone secures NVIDIA’s revenue visibility well into the latter half of the decade.

The Vera Rubin architecture ramp presents the next key product catalyst. NVIDIA confirmed the platform has entered full production, with CPU revenue projected to more than double in fiscal 2028 as both supply chains and customer demand strengthen. Historically, each new architectural generation has triggered a fresh wave of upgrade spending from NVIDIA’s largest enterprise clients.

Supply constraints remain the critical swing factor. Management indicated that current manufacturing capacity supports approximately 70% revenue growth, despite demand significantly exceeding that threshold. Bridging this gap—through expanded fabrication capacity or supply chain optimizations—could unlock growth rates surpassing current guidance.

Over the longer term, broadening NVIDIA’s customer base beyond a select group of hyperscalers into enterprises, AI-native startups, and sovereign AI initiatives could reduce reliance on any single client’s capital allocation decisions.

Track how these catalysts might move NVIDIA’s fair value (Free with TIKR) >>>

The most reliable approach is to examine the underlying data directly. TIKR provides complimentary access to institutional-grade financial datasets, enabling users to independently validate these projections.

Accessing NVDA’s profile reveals multi-year historical financials, consensus revenue and earnings forecasts, historical valuation multiple trends, and directional movements in analyst price targets.

Users can construct a complimentary watchlist to monitor NVDA alongside any other equities of interest. No credit card is required—only the transparent data necessary to form independent investment conclusions.

Disclaimer: Content published by TIKR is not intended as investment or financial advice, nor does it constitute a recommendation to purchase or sell securities. All analysis is derived from TIKR Terminal’s proprietary data and third-party analyst estimates. Coverage may not reflect the most recent corporate developments or material updates. TIKR holds no positions in any securities discussed. Thank you for reading, and best wishes for your investing journey!