Unconventional AI, a new hardware startup led by ex-Databricks AI chief Naveen Rao, has closed a US$475 million seed funding round, catapulting its valuation to US$4.5 billion. The company says the raise will accelerate development of proprietary AI acceleration hardware, part of a growing trend of startups building chips tailored for large AI models rather than relying solely on existing GPUs.
Backers reportedly include top-tier venture and growth investors betting on a future where custom silicon becomes a strategic advantage for AI training and inference workloads. With the new capital, Unconventional AI plans to expand its engineering team, ramp up chip prototyping, and begin pilot deployments with enterprise and research partners over the coming year.
Analysts see the funding as a strong endorsement of hardware-first approaches to AI, reflecting rising concerns over GPU shortages, rising costs, and energy consumption. If successful, Unconventional AI could challenge legacy vendors by offering more efficient, purpose-built solutions as demand for large models and compute-intensive tasks grows.
Surge AI CEO warns companies are prioritising ‘AI slop’ over solving real-world problems
