News & Trending Investing News

Nvidia Has an AI Weapon AMD Can't Match Right Now

Six million developers built a wall that no chip spec sheet can climb

Hand Holding Phone with NVIDIA (NVDA) Logo
Updated Sept. 1, 2026
Fact check checkmark icon Fact checked
Google Logo Add Us On Google info

Nvidia Corp. (NASDAQ:NVDA) reported $96.2 billion in fiscal Q2 2027 revenue on August 26, 2026. The number separating it from Advanced Micro Devices Inc. (NASDAQ:AMD) does not appear in any earnings release.

Six million developers locked into Nvidia's CUDA software ecosystem create switching costs AMD cannot replicate. A return on equity of 99.7% built on that foundation is among the clearest signs of financial success in semiconductors, and this is the competitive advantage behind the stock's margins.

Get a protection plan on all your appliances

Did you know if your air conditioner stops working, your homeowner’s insurance won’t cover it? Same with plumbing, electrical issues, appliances, and more.

Whether or not you’re a new homeowner, a home warranty from Choice Home Warranty could pick up the slack where insurance falls short and protect you against surprise expenses. If a covered system in your home breaks, you can call their hotline 24/7 to get it repaired.

For a limited time, you can get your first month free with a Single Payment home warranty plan.

Get a free quote

Nvidia's Q2 versus AMD's Q2 by the numbers

Nvidia posted $96.2 billion in fiscal Q2 2027 revenue, up 106% year over year, with data center revenue of $89 billion, up 117%, Zacks confirmed. AMD reported $11.5 billion in Q2 2026 revenue, up 50%.

  • Nvidia gross margins at 75.0% versus AMD's expected Q3 non-GAAP gross margin of 56%.
  • Nvidia Q3 guidance of $108 billion versus AMD Q3 guidance of $13 billion.
  • Nvidia's return on equity of 99.7% versus AMD's return on equity of 14.9%.

Both companies are growing their data center businesses, but Nvidia's revenue is roughly eight times larger and its profitability metrics are in a different category entirely.

CUDA's 18 years and six million developers as the real moat

CUDA has had 18 years of development and approximately six million developers building on its libraries, including cuDNN, TensorRT, and NCCL, MillionMiner reported. Every major AI framework optimizes for CUDA first, and the depth of answered questions, pre-built tools, and production-ready code gives it an ecosystem advantage AMD's ROCm cannot match.

Moving AI workloads from CUDA to AMD's ROCm incurs a real cost because teams must rewrite custom kernels, retune performance for a new architecture, and invest weeks of engineering effort. Nvidia's software locks customers in more effectively than any hardware specification, and your assessment of the stock's durability should account for that stickiness.

AMD's ROCm has closed the inference gap but trails on training

ROCm version 7 delivered significant inference speedups, and mainstream serving engines like vLLM run well on AMD hardware, as highlighted by MillionMiner. PyTorch and JAX are both supported as first-class frameworks, making AMD's inference story genuinely production-ready for specific workloads.

The gap remains on large-scale distributed training, where Nvidia's switched NVLink interconnect moves data between GPUs more efficiently than AMD's point-to-point fabric. Some high-performance kernels also remain CUDA-only, and independent testing shows AMD converting a lower fraction of theoretical peak performance into delivered work.

Nvidia at 23.5 times forward earnings versus AMD at 64.2 times

Nvidia trades at 23.51 times forward earnings compared with AMD's forward multiple of 64.21, making Nvidia the cheaper stock on a price-to-earnings basis despite its larger market capitalization, Zacks noted.

The valuation gap means you pay roughly 2.7 times more per dollar of AMD's future earnings than you pay for Nvidia's. AMD's faster revenue growth rate partially justifies the premium, but Nvidia's 75% gross margins and 99.7% ROE suggest the market may be underpricing Nvidia's earnings quality relative to AMD's.

Where AMD does hold an advantage over Nvidia

AMD's MI355X features 288GB of HBM3e memory compared to 180GB on Nvidia's B200, as reported by MillionMiner. A model that requires two Nvidia cards to hold its weights can sometimes fit on a single AMD card, reducing multi-GPU complexity and cloud rental costs by 15% to 40% on some workloads.

AMD also serves the second-source role for hyperscalers who want negotiating leverage against Nvidia pricing. Microsoft runs tens of thousands of AMD accelerators in Azure, some serving OpenAI workloads, and OpenAI signed a multi-year deal with AMD in late 2025. The second-source dynamic benefits AMD even when Nvidia remains the primary vendor.

Why the software moat matters more than the chip specs for your portfolio

Nvidia's 75% gross margins exist because CUDA gives it pricing power AMD does not have. Customers pay a premium for Nvidia hardware partly because switching to AMD costs engineering time and retraining, and that cost keeps Nvidia's margins elevated even as AMD's chips improve.

For your portfolio, the question is whether AMD's lower valuation on a forward P/E basis compensates for its significantly lower profitability and smaller competitive moat. Nvidia's Zacks Rank #2 Buy rating versus AMD's Rank #3 Hold reflects that difference in investment quality, and your allocation between the two should weigh moat durability alongside growth rates.

Risks of assuming Nvidia's dominance continues indefinitely

Custom silicon from Google, Amazon, and other hyperscalers represents a longer-term threat to both Nvidia and AMD. Google's TPUs and Amazon's Trainium chips bypass CUDA entirely, and their internal adoption reduces the addressable market for third-party GPUs over time.

Nvidia's CFO Colette Kress also warned on August 26 that gross margins will decline to 71% to 72% by the fourth quarter of fiscal 2027 as memory costs rise, CNBC reported. A margin decline of 300 to 400 basis points from current levels is meaningful, even when the revenue base is growing.

Bottom line

CUDA is the weapon AMD cannot match in 2026. Six million developers, 18 years of software depth, and switching costs that keep customers locked in drive Nvidia's 75% gross margins and 99.7% return on equity. AMD's chips compete on specs and even lead on memory capacity, but the software moat separates the two companies' financial profiles by a wide margin.

Determining when and how to start investing in one chipmaker over another depends less on which company ships the faster GPU and more on which software ecosystem your portfolio's growth thesis rests upon. Nvidia at 23.5 times forward earnings with a Zacks Rank #2 Buy and AMD at 64.2 times with a Rank #3 Hold give you a clear starting point for that comparison.

This article is for informational purposes only and should not be considered investment advice.

American Hartford Gold Benefits
  • American Hartford Gold helps individuals protect their retirement by rolling over IRAs and 401(k)s into physical gold.
  • Includes FREE IRA rollover and storage for up to 3 years.
  • Get up to $20,000 in free silver on qualifying purchases.


Financebuzz logo

Thanks for subscribing!

Please check your email to confirm your subscription.