Venture capital shifts toward deep tech startups building robots and chips

In short: Investors are putting more venture capital into deep tech startups, especially those building robots and computer chips.

What's going on

“Deep tech” is a broad label for startups based on hard science and engineering, not just new apps. It includes areas like robotics, semiconductors (computer chips), advanced materials, fusion, and quantum computing (a type of computing that uses unusual physics to solve some problems).

Recent programs and funding rounds show this interest is not just talk. Applied Ventures’ ASTRA 2024 program listed semiconductor and display technologies, plus robotics, and it selected robotics-focused startups such as Augmentus and CynLr.

Funding news in 2024 also pointed to big checks for robotics and chipmaking. Physical Intelligence reportedly raised $400 million to develop “foundation” AI models for robots (general-purpose AI, like a base model that can be adapted). On the chip side, Oso Semiconductor raised $5.2 million for phased-array chipsets (chips that help steer signals, like aiming a flashlight beam), and Positron raised $23.5 million for AI inference chips (chips that run AI after it is trained). DeepX also raised $80.5 million for on-device AI chip technology used in robotics and mobility.

What to watch

AI is pushing demand for the physical parts underneath it, like chips, sensors, and control systems. You can think of it like a delivery boom that requires more trucks and better roads, not just better maps. These companies can take longer and cost more to build than software startups, so watch whether investors keep funding them through slower timelines and whether more public programs and accelerators keep adding robotics and semiconductors to their cohorts.

Source: NYTimes