As generative AI continues to expand across industries, the demand for specialized AI chips is skyrocketing. Nvidia currently dominates the market, but its stronghold faces potential disruption from new entrants like Etched. Co-founded by Harvard dropouts Gavin Uberti and Chris Zhu, Etched aims to revolutionize AI hardware with its transformer-specific chip, Sohu. Unlike general-purpose GPUs, Sohu is an application-specific integrated circuit (ASIC) designed to run transformer models more efficiently. This focus allows Sohu to outperform traditional GPUs in speed and energy consumption. Etched has garnered significant investor interest, securing $120 million in a Series A funding round. The company plans to bring Sohu to mass market soon, betting on its unique design to meet rising energy and performance demands in AI. While competition looms and the future of transformer models remains uncertain, Etched’s innovative approach and early customer reservations suggest a promising start.
Nvidia Faces New Challenger in AI Chip Market – Meet Etched’s Sohu
Etched’s Sohu chip aims to outperform Nvidia GPUs by focusing solely on transformer models.
1–2 minutes
TOP STORIES
U.S. proposes a new AI incident notification system to enhance national security discussions with China …
The debate over AI safety intensifies as industry leaders clash on regulation and innovation …
Trump labels AI safety concerns as hoaxes and plans to form an AI Force …
Google’s Gemini conducted autonomous hacks on three companies during tests …
AI misjudgment nearly led to a military conflict with China this spring …
Hackers successfully exploited OpenAI’s vulnerabilities using Anthropic’s Claude model, prompting urgent concerns in AI security …
