ADAS & Autonomy Stacks
ADAS and autonomy stacks span in-house OEM software, merchant perception and compute vendors, safety-critical operating systems, Tier 1 integrators and robotaxi or autonomous-trucking developers. Recent activity centres on XPeng's push to license its stack and launch robotaxi registration, Tesla's regulatory friction in Europe and the US, Aurora's driverless-truck scale-up targets, and strong automotive software revenue at BlackBerry's QNX unit. The main structural tension is that large OEMs keep bringing ADAS software in-house, which pressures merchant suppliers such as Mobileye while concentrating demand on a small group of silicon and platform providers.
Active institutional exposure flipped from a net outflow of $4.2 billion in Q1 2026 to a net inflow of $11.6 billion in Q2 2026. 66.7% of companies in this sub-theme saw increased active positions in Q2 2026.
Overview
ADAS and autonomy stacks are the software, compute and sensing layers that let vehicles assist or replace a human driver. The sub-theme has several distinct layers. Vertically integrated OEMs develop their own stacks, including XPeng with its Turing AI chip and Tesla with internal SoC design teams. Merchant ADAS suppliers such as Mobileye and Ambarella sell silicon and perception software, and Mobileye's EyeQ platform reaches the market largely through Tier 1 customers including Aptiv, Valeo and ZF. Nvidia sits across the stack as a compute supplier to XPeng, Pony AI and Magna's planned DRIVE AGX Thor-based ADAS and autonomous driving solutions, while also competing with Mobileye and Ambarella. BlackBerry's QNX provides safety-certified operating systems on which developers such as WeRide have built L2++ ADAS platforms. Robotaxi and autonomous-trucking developers form a separate layer. Pony AI, WeRide and Baidu's Apollo Go reach riders through ride-hailing platforms such as Uber and Lyft. Aurora works with OEM and fleet partners including Toyota, PACCAR, FedEx, Werner and Schneider National. Semiconductor and component makers such as NXP, STMicroelectronics, Qualcomm and Aptiv supply the hardware and in-cabin sensing around these stacks, and contract manufacturers such as Magna build the vehicles. Value accrues differently by layer: licensing and per-unit royalties for merchant stacks, vehicle sales for integrated OEMs, and per-mile or service fees for autonomous mobility and freight.
Investment Thesis
Investment thesis details are part of the Pro plan.
Bottleneck Tracker
Bottleneck details are part of the Pro plan.
Key Companies
Related Events
Factory-installed Waymo hardware on Toyota's line replaces the retrofit model of buying finished cars, shipping them and adding sensors, wiring and compute. That could lower per-vehicle cost and speed fleet scaling for Waymo. It also positions Toyota as a high-volume OEM supplier of purpose-built robotaxis, adding to Waymo's Jaguar I-Pace, Zeekr-based Ojai and Hyundai Ioniq 5 platforms. Japan and UK expansion extends Waymo beyond the US and raises competitive pressure on Tesla, Baidu Apollo Go, Wayve and Uber-linked robotaxi partners. Toyota's work on personally owned autonomous vehicles and the Woven telemetry tie-up show the partnership may extend beyond robotaxis, but commercial timelines and unit volumes are not disclosed.
A $3.5 billion formal deal monetizes the previously noted IP toll structure between Nvidia and MediaTek, locking in a multi-year revenue stream for both parties and cementing MediaTek as a primary channel for Nvidia AI IP into non-datacenter markets. This structurally disadvantages Qualcomm and other SoC vendors competing for PC and automotive AI silicon. MediaTek gains engineering and brand credibility for enterprise AI chip customers who previously defaulted to Nvidia direct. Automotive OEMs selecting MediaTek-Nvidia combined silicon get a validated ADAS and in-cabin AI stack, accelerating their design-win cycles. The deal deepens Nvidia's AI IP licensing revenue diversification beyond its own GPU hardware.
This deal structurally cements Mobileye's position as a preferred ADAS supplier for a major global OEM at a time when competition from in-house stacks (Tesla) and rivals (Qualcomm, Nvidia) is intensifying. For Stellantis, which has been cutting costs and restructuring, outsourcing ADAS to Mobileye rather than developing in-house is a capital-efficient path to regulatory compliance and consumer feature parity. The 2027 production timeline creates a multi-year revenue visibility improvement for Mobileye. Rival ADAS suppliers — including Qualcomm's Snapdragon Ride and Continental — lose a potential Stellantis socket. The deal also signals that mid-tier OEMs under financial pressure will increasingly rely on third-party ADAS platforms rather than vertical integration.
Removing the federal requirement for human-operable controls is a structural rule change that unlocks the full commercial deployment of purpose-built robotaxi and autonomous fleet vehicles in the US. Companies with driverless-first designs gain a decisive regulatory advantage over competitors still building around hybrid human-AI control architectures. This accelerates the timeline for robotaxi fleet operators to deploy at scale without costly retrofits. Traditional auto suppliers dependent on pedal, steering column, and driver interface components face a long-run structural decline in addressable market. Lidar, sensor, and autonomy stack providers benefit from expanded AV deployment opportunity.