Etched has announced a $120 million funding round to develop its specialized AI chip, Sohu, which aims to transform AI processing by embedding the transformer architecture directly into its hardware. This approach promises to deliver unprecedented performance for transformer models, positioning Etched as a formidable competitor to industry giant Nvidia. Etched’s founders, who dropped out of Harvard, have garnered substantial backing from notable investors including Peter Thiel. The company is betting on the scaling laws of AI, asserting that larger models and better training data will lead to superintelligent systems. By focusing on transformers, Etched claims that its Sohu chip will be significantly faster and more cost-effective than Nvidia’s next-gen GPUs. The company plans to partner with TSMC for production and has already secured reservations from major AI companies. Etched’s strategy underscores the inevitability of specialized chips in the AI market, predicting that their adoption will revolutionize AI product economics and capabilities.

Source.

TOP STORIES

Samsung's Bid to Challenge TSMC's Chip Manufacturing Dominance
Google is partnering with Samsung to produce a new TPU, but TSMC remains crucial …
Attorneys Must Face the Consequences of AI Hallucinations
Attorneys can no longer claim ignorance of AI hallucinations as courts demand accountability …
Anthropic's AI Access Suspension Sparks Debate in India's Tech Sector
Anthropic’s suspension of AI model access highlights India’s reliance on foreign technology and sparks discussions on developing domestic AI capabilities …
The Quantum Revolution - Transforming Technology and Security
Quantum computing is transforming industries, but it poses significant cybersecurity risks …
Investigation Launched Into OpenAI by State Attorneys General
A coalition of state attorneys general has opened an investigation into OpenAI …
Anthropic Faces AI Export Controls - A New Era of Regulation
The U.S. government’s export control directive has forced Anthropic to disable its new AI models, raising questions about regulation and …

latest stories