HardwareJul 13, 2026, 12:40 PM

Korean AI Startup, Furiosa AI, Is Doubling Its Chip Production To 50,000 Units Next Year While Its Upcoming 2nm “Stork” Chip Challenges NVIDIA With The “World’s Best Inference” - Wccftech

30-second summary

South Korean AI chip startup Furiosa AI will double its production to 50,000 units in 2025 and claims its upcoming 2nm Stork chip delivers superior inference performance compared to NVIDIA.

TickrWire
Key takeaways
  • Furiosa AI plans to double chip production to 50,000 units in 2025, signaling aggressive expansion in the AI hardware market.
  • The company's 2nm "Stork" chip is positioned as a high-performance inference solution, directly competing with NVIDIA's offerings.
  • Furiosa AI targets the inference segment, where demand is growing rapidly due to the rise of generative AI and LLMs.
  • The startup's claims about "world's best inference" performance remain to be validated by independent benchmarks.
Full story

Furiosa AI, a South Korean AI chip startup, announced plans to double its chip production capacity to 50,000 units annually starting in 2025. The company is positioning itself as a serious contender in the AI hardware space with its upcoming 2nm process node chip, codenamed "Stork." Furiosa AI claims the Stork chip will deliver "the world's best inference" performance, directly challenging NVIDIA's dominance in AI accelerator chips.

The move comes as global demand for AI inference chips surges, driven by the rapid adoption of large language models and generative AI applications. Furiosa AI's strategy focuses on high-performance, energy-efficient chips tailored for inference workloads, a segment where NVIDIA currently holds a near-monopoly. The company's production expansion and advanced process technology suggest it aims to carve out a significant market share in the AI chip ecosystem.

Industry analysts note that while Furiosa AI's claims are ambitious, the company's progress will depend on its ability to deliver on performance benchmarks and secure partnerships with major AI infrastructure providers.

Why this matters
Developers

Opportunity to evaluate Furiosa AI's Stork chip for inference workloads and compare it with existing solutions.

Businesses

Potential alternative to NVIDIA for AI inference hardware, offering a new vendor in a highly concentrated market.

Investors

Signals growing competition in the AI chip space, with implications for market dynamics and investment opportunities.

Everyone

Highlights the global race to develop advanced AI hardware beyond NVIDIA's current dominance.

Glossary
inference
The process of running a trained AI model to make predictions or generate outputs, as opposed to training the model.
2nm process node
A semiconductor manufacturing technology node representing the size of transistors, where smaller numbers indicate more advanced and efficient chips.
Sources · 1
Read next
More stories
TickrWireAI News Intelligence

We aggregate, verify, summarise and explain the latest artificial intelligence news from open, legal sources.

Daily AI digest

Top AI stories, summarised, in your inbox each morning.

© 2026 TickrWire. Summaries and analysis are AI-generated and may contain errors.