Autonomous Vehicle Physical AI
Companies applying large-scale neural Physical AI architectures — including end-to-end learned driving models and foundation models for autonomy — to self-driving cars, trucks, and air taxis.
CAPITAL FIGURES ARE MEDIA-EXTRACTED ESTIMATES, NOT VERIFIED FILINGS.
EXTRACTED FROM 25+ PODCASTS & VC NEWSLETTERS · MEDIA-REPORTED FIGURES, NOT VERIFIED FILINGS
Nvidia is the compute and capital spine of autonomy
Nvidia appears in 50 deals in this theme, far ahead of Blackstone (8), AMD Ventures (5), Bain Capital (4) and BlackRock (4). It is also a Wayve investor and partner, and it is extending itself from chips into models and platforms: the open Nemotron 3.5 Super VL, the GR00T N1.7-3B VLA foundation model, and corporate builds on Nemotron such as Salesforce's. Jensen Huang credits Tesla with turning automotive into a computing industry, and Nvidia hardware such as the RTX 5090 is now the default onboard and workstation compute in physical-AI research. Private credit lenders like Blackstone and Apollo also back Nvidia-linked buildouts. The result is that capital, silicon and foundation-model tooling for autonomy are increasingly sourced from one ecosystem.
Wayve's vehicle-agnostic, end-to-end AI Driver runs without HD maps or custom hardware and is sold to automakers and fleets. Its partners and investors include Nissan, Mercedes-Benz, Stellantis, Uber, Microsoft and Nvidia, and it builds GAIA generative world models for training and validation. Momenta, Mobileye and Applied Intuition (simulation and validation tooling) are positioned around the same shift from rules-based stacks to learned models. Public sentiment is following: investors now describe Tesla FSD and Waymo as proof that autonomy 'has arrived' and warn that skepticism based on the 2017 experience is a trap.
Why it matters · Software licensing across many OEMs scales faster than owning fleets, and it favors the few companies with the best end-to-end models and data.
Moove AV owns and operates Waymo's all-electric fleets in Phoenix, Miami and London, and plans to finance up to 6,000 Waymo vehicles with a $1.2B debt raise. Uber is the lead investor in Moove AV and acts as the aggregation layer for AV partners including Nuro/Lucid, Pony.ai and WeRide. Large non-equity financings, such as the $5B round with Sixth Street, Blackstone and PIMCO, show infrastructure and credit investors entering. Technology developers can scale the driver while asset-heavy operators carry the balance-sheet risk.
Why it matters · Asset-backed debt lets robotaxi networks scale without diluting AV developers, and it opens a new class of infrastructure-style investments.
Weekly capital swung from $10.7B across 7 deals (week of 7/27) to $237M across 6 deals (week of 10/5), then back to $6.2B across 3 deals (week of 10/12). Two $600M Series C signals at a $3.5B valuation and a $5B debt-style financing drive the recent spikes. In 90 days, 'unknown' stage deals account for $33.6B and seed for $5.3B across only 4 deals, so capital is concentrated in a few outsized checks rather than broad deal flow.
Why it matters · Headline totals overstate breadth, so investors should expect a thin set of well-capitalized leaders and a harder environment for sub-scale entrants.
Siemens, Hitachi Ventures and Indra are leading deep-tech rounds, and Nvidia is itself investing in French startups and making a reported $30B acquisition of Reflection. The 'vertical integration out of SpaceX and Tesla' ethos is now cited as the template for new hardware-plus-AI companies. Autonomy players in trucking, air taxis and delivery (Waabi, Torc, Gatik, Joby) must decide between partnering with a platform and building the full stack.
Why it matters · Corporate strategics are pricing the market, so independent AV startups face rising pressure to align with a platform or be acquired.