Meta and xAI releases bolster Nvidia's moat, says Gabelli
Gabelli Funds' John Belton argues that Meta's Muse Spark 1.1 and xAI's Grok 4.5 validate Nvidia's infrastructure dominance, countering fears about custom AI chips. He notes hyperscalers like Meta provide near-term spending visibility, accounting for 50% of Nvidia's business, while competition among AI labs benefits the chipmaker. BofA previously projected a 78% upside for Nvidia, citing a 65-70%+ market share through 2030.

*this image is generated using AI for illustrative purposes only.
Gabelli Funds portfolio manager John Belton says recent model releases from Meta Platforms Inc. and Space Exploration Technologies Corp’s xAI suggest Nvidia Corp.’s competitive moat remains firmly intact despite rising fears over custom silicon. Belton pointed to Meta’s Muse Spark 1.1 and xAI’s Grok 4.5 as evidence that frontier AI developers continue to rely on Nvidia’s infrastructure to train their most advanced models. He noted that both models were trained on Nvidia infrastructure, indicating a clear value proposition for the chipmaker's stack.
The AI Race Runs on Nvidia
The observation addresses investor concerns regarding how long Nvidia can maintain leadership with hyperscalers including Meta, Alphabet Inc., Amazon.com Inc., and Microsoft Corp. investing heavily in custom AI chips. While in-house silicon efforts expand, Belton argues the latest generation of frontier models shows Nvidia remains the platform of choice for the industry’s most demanding AI workloads. He added that fragmentation in the large language model space is beneficial for Nvidia, as a winner-take-all market would be less attractive over the long run.
Hyperscaler Spending Visibility
Belton also highlighted reports suggesting Meta’s AI infrastructure spending in 2027 could exceed Wall Street expectations. This reinforces the view that the hyperscaler capital expenditure cycle is far from over. Hyperscalers account for roughly 50% of Nvidia’s business, according to Belton. While the market has raised concerns about revenue durability due to hyperscalers operating around break-even free cash flow, Meta’s expanding ambitions provide greater near-term visibility into AI spending.
Peer Valuation Context
Separately, Bank of America has previously reiterated a Buy rating on Nvidia with a price target of $350, citing a 78% upside potential. The firm noted Nvidia trades at a valuation discount to mega-cap peers despite forecasting gross margins in the mid-70% range. BofA expects Nvidia to maintain a 65-70%+ share of the global AI accelerator market through calendar year 2030.
| Company | 2026 P/E ratio | 2027 P/E ratio | 2028 P/E ratio |
|---|---|---|---|
| Nvidia | 22.9x | 15.7x | 12.4x |
| Apple | 35.0x | 30.9x | 28.5x |
| Microsoft | 21.8x | 18.7x | 15.5x |
| Alphabet | 25.6x | 24.1x | 20.0x |
| Amazon.com | 24.0x | 21.4x | 17.3x |
| Meta Platforms | 15.9x | 16.3x | 13.4x |
| Average without NVDA | 24.5x | 22.3x | 19.0x |
At what point might the efficiency gains from custom hyperscaler silicon outweigh the benefits of Nvidia's established ecosystem?
How will Nvidia's revenue mix evolve if hyperscalers successfully transition their inference workloads to in-house chips while retaining Nvidia only for training?
Could the fragmentation of the LLM market eventually lead to specialized hardware requirements that custom silicon is better positioned to address than Nvidia's general-purpose GPUs?

































