Nvidia's Market Dominance Erodes as Competitors Close In Post-Huang
With Jensen Huang's departure, Nvidia faces mounting pressure from AMD, cloud giants' custom silicon, and specialized AI chip startups—even as the company maintains an 80% accelerator market share.
Last verified:
The Leadership Transition and Market Timing
Nvidia’s departure of CEO Jensen Huang coincides with a period of intensifying competitive pressure in the AI accelerator market. According to Wired AI, CFO Colette Kress has assumed leadership as the company faces mounting challenges from rivals. The timing presents a strategic vulnerability: as an unnamed AI company executive told Wired AI on condition of anonymity, “everything sold today was designed two-to-three years ago,” meaning Nvidia’s current and near-term product lineup reflects decisions made before many of today’s competitive threats materialized. Kress must now chart the company’s technological direction for the 2028–2029 timeframe while competitors make their own advances.
AMD Gains Meaningful Traction
AMD has transformed from a marginal player to a credible alternative in AI infrastructure. According to Wired AI, citing Silicon Analysis data, AMD’s share of the AI accelerator market grew from 2% in 2022 to between 7% and 10% by 2026—a significant acceleration in adoption. While Nvidia retains 80% of the market, AMD’s upward trajectory reflects years of concerted investment in competitive AI chips and suggests the dominance picture is shifting, albeit slowly.
Cloud Giants and Specialized Startups Challenge GPU Hegemony
The erosion of Nvidia’s advantage extends beyond traditional semiconductor rivals. According to Wired AI, Google, Amazon, and Microsoft are each developing specialized silicon for AI model training and inference, motivated by cost reduction and supply-chain independence. These cloud providers have captured market gains similar in scale to AMD’s gains, though the article does not quantify their combined share.
Startups have entered the arena with alternative approaches. Wired AI reports that Cerebras manufactures large-scale chips optimized for training, while AMD acquired Taalas, a company that burns model weights directly into silicon to improve computational efficiency. However, Wired AI notes that adopting these non-Nvidia alternatives requires significant software rewrites, creating switching costs that protect Nvidia’s existing customer base.
Inference and CPU Markets: Nvidia’s Blind Spots
Patrick Moorhead, chief analyst at MoorInStrat, argues that the rise of AI agents makes inference—running already-trained models—increasingly important to customer workloads, according to Wired AI. In this domain, Nvidia has a narrower advantage compared to cloud hyperscalers and startups. Nvidia did acquire Grok, a startup specializing in inference chips, in December 2025, but the company historically has underinvested in CPU-class processors—an area where demand is growing alongside AI assistant deployments, Wired AI reports.
Why This Matters
Nvidia’s strategic moment is constrained by product-development cycles that extend two to three years. While the company maintains an 80% share of the accelerator market—a formidable moat—the trajectory of competitive gains, combined with leadership transition risk and Huang’s departure as the visionary architect of Nvidia’s AI-era dominance, raises questions about whether Kress can maintain the company’s technological lead as custom silicon, AMD’s advances, and specialized startups capture use-case-specific demand. Teams evaluating long-term GPU sourcing, particularly for inference workloads, now have credible alternatives worth re-evaluating as costs and architectural fit improve over the next 18–24 months.
Frequently Asked Questions
What happened to Jensen Huang at Nvidia?
According to Wired AI, Huang has departed from his role as Nvidia CEO. The article does not detail the specific circumstances, but notes the timing is challenging given competitive pressures.
How much market share has AMD gained in AI accelerators?
AMD's share of the AI accelerator market has grown from 2% in 2022 to between 7% and 10% in 2026, according to Wired AI reporting.
Why are cloud providers building their own chips?
Google, Amazon, and Microsoft are developing custom silicon to reduce costs and gain independence from Nvidia's supply chain, Wired AI reports.
What is Nvidia's remaining competitive advantage?
Nvidia still controls 80% of AI accelerator sales and has invested heavily in software ecosystems and research funding, though inference and CPU markets present weaker positions.