KOSUNI CRYPTO
Korea's Crypto Pulse, in English
$BTC $ETH $XRP $SOL $DOGE
Market Analysis
Bullish

Nvidia vs. OpenAI Showdown... Semiconductors Benefit from Whale Battle

Published September 25, 2026 6:01 PM · 7 views $NVDA $HUGG
Nvidia vs. OpenAI Showdown... Semiconductors Benefit from Whale Battle

Nvidia and OpenAI, two titans of the AI industry, have simultaneously made major moves. On August 3 (Eastern Time), Nvidia announced it would acquire Hugging Face, the world's largest open-source AI platform, for $12.9 billion (approximately 17.5 trillion won). This follows last year's acquisition of Groq, an inference-focused AI chip design company, for $20 billion.

Hours later, OpenAI unveiled its latest flagship model, GPT-6 'Astra,' with a limited release. The model demonstrates significant improvements in ARC-AGI-3 capabilities, rising from 7.8% to 99.9%, and uses only 20% of the tokens required by Anthropic's top-tier models for coding tasks. OpenAI CEO Greg Brockman welcomed users into the 'AGI era,' stating, 'Welcome to the age of AGI.'

Both open-source AI and cutting-edge closed-source AI advancements have spurred market reactions. On August 4, semiconductor giants Samsung Electronics and SK Hynix led gains in South Korea's stock market, with data center and sovereign AI-related stocks like SK Telecom and Samsung SDS also surging.

In the U.S., despite broader market declines due to rising interest rates following strong August employment figures, memory and semiconductor hardware companies—including Micron, Sandisk, Western Digital—rose significantly. Meanwhile, hyperscalers and custom chip leaders like Broadcom underperformed. The reason? Nvidia's acquisition of Hugging Face aligns with its strategy to dominate the AI ecosystem through open models.

Hugging Face serves as a platform where developers upload and download AI models, similar to GitHub for code. With 18 million developers, 200,000 enterprises, 3 million models, 500,000 datasets, and 1 million applications, it generates around $150 million in annual revenue. Nvidia paid 86 times its valuation—$12.9 billion—to secure this platform.

Nvidia's strategy is to strengthen its ecosystem beyond GPUs by leveraging platforms like Hugging Face, which could serve as the 'CUDA' of AI models. CEO Jensen Huang has long advocated for open-weight models, stating they expand AI access and bolster U.S. technological leadership. Nvidia plans to maintain Hugging Face as an open platform without forcing chip usage.

Analysts view the acquisition as a hedge against growing competition from closed-model builders like OpenAI, Anthropic, and Google, who are developing their own chips. Raymond James described it as 'a hedge against the trend of closed-model builders running workloads on custom chips.' Barron's called it a 'cheap insurance policy' for Nvidia, which has a market cap of $5 trillion.

OpenAI's GPT-6 'Astra,' trained on over 100,000 Nvidia GPUs at Texas StarGate data centers, highlights the scaling law in AI. This model emphasizes efficiency and cost-effectiveness, with CEO Sam Altman stating, 'The goal is the best price-to-performance at every level of intelligence.' The launch has eased concerns about token price declines due to open models, as Goldman Sachs noted that improved capabilities reignite demand for tokens.

Market reactions reflect renewed confidence in AI's expansion. Hardware stocks surged as investors bet on increased compute and memory needs regardless of whether models are open or closed. While companies like Broadcom (focused on custom chips) lagged, memory and data center infrastructure firms saw gains. UBS highlighted that hardware roadmaps are shifting toward memory capacity, bandwidth, and interconnects.

Korean Source

This article is an English localization of a Korean-language crypto news report. Original headline: 엔비디아·오픈AI 격돌…고래싸움에 반도체 웃었다 [빈난새의 빈틈없이마켓]