Monday, November 16, 2026

AI Industry Ecosystem Grows More Multipolar ... NVIDIA's Dominant Position Is Under Threat

Input
2026-11-16 14:23:58
Updated
2026-11-16 14:23:58
NVIDIA CEO Jensen Huang resigned on the 12th (local time), putting key decisions at risk of being left in limbo. The photo shows former CEO Huang at a celebration of the U.S.-China summit held in Beijing, China, on May 14. AP/Newsis

[Financial News] Amid the turmoil of the AI industry's ecosystem reshuffle, the leadership vacuum left by the resignation of NVIDIA CEO Jensen Huang has further accelerated the arrival of a “multipolar era” in the AI industry. 
According to industry sources and others on the 16th, as customers accelerate the manufacturing of their own semiconductor chips and competition in the AI industry shifts from securing graphics processing units (GPUs) to building computing ecosystems, the erosion of NVIDIA's monopoly is also accelerating.
 
AI industry splinters into multiple ecosystems and grows more multipolar ... NVIDIA's contract volume is halved

NVIDIA's dominant position is already fragmenting across multiple ecosystems and becoming more multipolar. JP Morgan said on the 14th (local time; all times hereafter are local) that NVIDIA accounted for 39% of future contracts, measured by power capacity (GW). Broadcom (24%), Google's TPU (16%) and AMD (15%) followed. The analysis is that the share of the dominant market leader, which controlled 85.2% of the AI accelerator market in the second quarter of 2025, will fall to half or less. 
Jensen Huang's resignation has heightened uncertainty by leaving no one to make decisions on pressing challenges: technical and managerial issues, as well as redefining relationships with customers. Decisions and responses to the reshaping of the AI ecosystem are likely to be left in limbo, and customer confidence is also wavering.
Customers have been seeking alternative suppliers rather than risk relying on a single company, NVIDIA, for semiconductor chips, and their efforts to diversify supply chains are accelerating. The share of contracts for customized semiconductors has already neared 46%.
The movement was also reflected in stock market fluctuations. On the 13th, when the Nasdaq index fell 0.4% and NVIDIA shares dropped 4.34%, NVIDIA rivals AMD and Broadcom rose 5.54% and 5.86%, respectively.
AMD's market capitalization stood at $736.5 billion as recently as the 6th, but surged to $960 billion (about 1,280 trillion won) in just four business days. Its market-cap ranking also jumped three places, from 19th to 16th.

The fundamental threat is customers making their own chips ... the ecosystem is also being reshaped

Anna Woo of Floodlines Research said that direct competitors AMD and Broadcom, along with Alphabet (Google) in the tensor processing unit (TPU) field, “stand to benefit in the short term.” In AI accelerators, AMD overlaps with NVIDIA in more areas than any other competitor, including general-purpose graphics processing units (GPUs) and AI server racks, while Broadcom is strong in custom chips (ASICs) and networking. Cerebras, which has wafer-scale engine (WSE) technology that turns a single giant wafer into one chip, also saw its stock jump more than 17% that day on better-than-expected results and a new contract with Microsoft.
The pursuit by AMD and others is not the only thing shaking NVIDIA's dominant position. There is a more fundamental threat. Customers are becoming competitors by making their own semiconductor chips, and are reshaping the ecosystem as well. Google, Amazon, Microsoft and Meta, among others, are accelerating the development of their own chips, backed by data centers and software. Cloud providers are managing costs and performance by leveraging their own semiconductors. 
These companies, including Google, are stepping up efforts to build integrated systems that can replace the platform ecosystem NVIDIA has spent more than 20 years building.
Model companies such as Anthropic and OpenAI are diversifying their supply chains by using different types of chips from multiple companies. The strategy is clear in the IPO prospectus Anthropic submitted to the U.S. Securities and Exchange Commission (SEC) on the 15th. Anthropic said it would use four types of chips: Amazon Trainium, Google's tensor processing units (TPUs), NVIDIA GPUs and AMD GPUs. TPUs and Trainium are leading examples of ASICs. 

Anthropic to use four types of chips, including Amazon Trainium, Google TPU, NVIDIA GPU and AMD GPU

Anthropic will use Amazon's own AI chips, including Trainium, to secure up to 5 gigawatts (GW) of computing capacity under its contract with Amazon. Anthropic's computing strategy is to choose chips with favorable price and performance characteristics for each type of task. It is targeting both favorable price negotiations with suppliers and tailored performance. 
Instead of assigning all computing to a single expensive general-purpose GPU, as in the past, hardware is now assembled from in-house chips or company-specific custom accelerators to suit tasks such as training and inference. As a result, the ASIC segment has also grown. The global ASIC market is projected to expand from $22.78 billion last year to $24.21 billion this year and $41.68 billion in 2034, with a compound annual growth rate (CAGR) in the 7% range.
Cloud providers are using their own chips to narrow the technology gap with NVIDIA and strengthen their bargaining power over NVIDIA GPUs. As they expand their own semiconductors and AI ecosystems, NVIDIA faces not only market-share erosion but also weakening pricing power.
Anthropic CEO Dario Amodei's remark that “securing computing resources is the biggest business dilemma the company faces” shows the company's determination to diversify its computing supply chain. In remarks made public on the 15th, he said, “If you have too few computing resources, it's difficult to develop new technologies; if you have too many, the financial burden is substantial,” signaling an intention to manage resources economically. 
Experts said that NVIDIA's key challenge—and the key to maintaining its dominance—is how to sustain medium- and long-term cooperative relationships with its customers.

[email protected] Lee Seok-woo, International Affairs Specialist Reporter