Nvidia has reportedly reduced the number of approved AI chip buyers in parts of Asia by more than half following the implementation of stricter compliance measures designed to prevent its advanced processors from reaching China. The company created a new "white list" of authorized buyers, resulting in the exclusion of many smaller neo-cloud providers that rent AI computing capacity, according to a report by Reuters. Nvidia has spent the past several months strengthening due diligence efforts in Singapore, Malaysia, and Japan to ensure adherence to U.S. export controls aimed at restricting China's access to cutting-edge AI hardware.

Compliance and Reapplication Process

The tighter screening process specifically targets the diversion of advanced AI chips to Chinese-linked organizations through third countries. While more than half of Nvidia's previous customers did not make the initial approved list, the report indicates that excluded companies can update their compliance procedures and reapply for approval. Nvidia did not immediately respond to requests for comments regarding these changes.

Regulatory Context and Market Impact

The reported changes align with the U.S. Commerce Department's issuance of new guidance in May, warning companies against allowing advanced AI chips to reach overseas subsidiaries of Chinese firms. Officials have raised concerns that Nvidia's latest Blackwell AI processors could be diverted to Chinese-linked organizations through countries like Malaysia. Meanwhile, Chinese officials have reportedly informed Alibaba Group Holding Ltd., ByteDance, and DeepSeek that they could soon be cleared to purchase certain Nvidia H200 AI chips.

Nvidia shares closed Monday down 3.52% at $203.53 and slipped another 0.14% to $203.25 in after-hours trading.

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.