Pony.ai Announces New Generation Autonomous Driving Compute Platform Built on NVIDIA DRIVE Hyperion

Pony.ai (NASDAQ: PONY) unveiled a next‑generation autonomous driving domain controller built on NVIDIA DRIVE Hyperion and powered by DRIVE AGX Thor with NVLink, targeting L4 robotaxis and broader autonomous mobility customers.

Rhea-AI Summary

Pony.ai (NASDAQ: PONY) unveiled a next‑generation autonomous driving domain controller built on NVIDIA DRIVE Hyperion and powered by DRIVE AGX Thor with NVLink, targeting L4 robotaxis and broader autonomous mobility customers.

The platform supports single‑chip and multi‑chip configurations with up to 4000 FP4 TFLOPS combined compute, improved energy efficiency, redundancy, and flexible cooling and deployment options. Pony.ai reported >500% YoY shipments of its Fangzai controller in 2025 and says it reached unit‑economics breakeven in two major Chinese markets, targeting >3,000 robotaxis and >20 cities by end of 2026.

Positive

• Platform delivers up to 4000 FP4 TFLOPS combined compute
• Fangzai shipments rose 500%+ YoY in 2025
• Reached unit‑economics breakeven in two Chinese markets
• Supports single‑chip and multi‑chip configs with NVLink
• Targets >b>3,000 robotaxis and >b>20 cities by end‑2026

Negative

• None.

Key Figures

Compute performance: 4000 FP4 TFLOPS Shipment growth: More than 500% YoY Robotaxi fleet target: More than 3,000 vehicles +5 more

Compute performance

4000 FP4 TFLOPS

Combined maximum performance of dual NVIDIA DRIVE Thor SoCs

Shipment growth

More than 500% YoY

Fangzai domain controller shipments in 2025

Robotaxi fleet target

More than 3,000 vehicles

Planned robotaxi fleet by end of 2026

City footprint target

More than 20 cities

Planned global robotaxi footprint by end of 2026

SoC count

Four NVIDIA DRIVE AGX Orin SoCs

World’s first L4 Robotaxi domain controller in mass production in 2025

Collaboration start year

2017

Start of long-standing collaboration between Pony.ai and NVIDIA

Gen-7 robotaxis

Seventh-generation

Current robotaxi generation powered by earlier Orin-based controller

Deployment markets

Dozens of countries

Customer base for Pony.ai’s automotive-grade domain controllers

Previous AI Reports

5 past events · Latest: Apr 22

Same Type 5 events

1. Apr 22

   AGM announcement

   24h Move

   -2.1%

   Announcement of AGM date, record dates, and reference to filed Form 20-F.

2. Apr 22

   Annual report & ESG

   24h Move

   -2.1%

   Filing of 2025 Form 20-F and inaugural ESG report for investors.

3. Apr 10

   AI engine upgrade

   24h Move

   +2.0%

   Launch of PonyWorld 2.0 to enhance L4 safety, comfort, and efficiency.

4. Apr 8

   Europe robotaxi launch

   24h Move

   +11.1%

   Launch of Europe’s first commercial robotaxi service in Zagreb with partners.

5. Apr 7

   Singapore AV progress

   24h Move

   -2.0%

   Regulatory approval for by-invite rides on a defined Singapore AV route.

24h Move is the share-price change in the day after each event; other market factors may also have contributed.

Key Terms

l4 autonomous driving, nvidia drive hyperion, nvidia drive agx thor, nvidia nvlink, +4 more

8 terms

l4 autonomous driving technical

"designed for both Pony.ai's L4 autonomous driving platform and a broader set"

Level 4 (L4) autonomous driving describes vehicles that can perform all driving tasks and handle road conditions on their own within a limited area or set of scenarios, without requiring a human to steer, brake or monitor continuously; a human can still take over if allowed. Investors care because L4 requires large up-front investment in sensors, software and testing but can open new predictable revenue streams—like robotaxis or automated delivery—while also exposing companies to regulatory, safety and capital risks.

nvidia drive hyperion technical

"the new controller is built on the NVIDIA DRIVE Hyperion platform and powered"

NVIDIA Drive Hyperion is a complete development and validation kit that combines sensors (cameras, radar, lidar), computing hardware, and driving software used to build and test self‑driving vehicle systems. For investors, it matters because such platforms act like a standardized laboratory or toolkit that lets carmakers and suppliers prototype, certify and scale autonomous features faster and with lower technical risk—potentially accelerating product rollout and revenue opportunities.

nvidia drive agx thor technical

"powered by NVIDIA DRIVE AGX Thor with NVIDIA NVLink, supporting Pony.ai's next"

A high-performance automotive computer system that combines specialized chips and software to run the artificial intelligence needed for advanced driver assistance and self-driving features. Think of it like the brain and nervous system of a car that processes sensor data (cameras, radar, lidar) in real time so the vehicle can see, decide and act; investors watch it because adoption and licensing of such platforms can drive large, recurring revenue and determine competitive leadership in the auto tech market.

nvidia nvlink technical

"powered by NVIDIA DRIVE AGX Thor with NVIDIA NVLink, supporting Pony.ai's next"

NVLink is NVIDIA’s high-speed chip-to-chip connection that lets graphics processors and other chips share large amounts of data much faster than standard wiring. Think of it as a dedicated express lane between processors that reduces delays and lets multiple chips act like a single, more powerful engine; for investors, that can translate into stronger performance for data-center and AI workloads, greater product differentiation, and potential revenue and margin benefits.

system-on-a-chips (socs) technical

"between the two DRIVE Thor system-on-a-chips (SoCs), achieving a combined"

A system-on-a-chip (SoC) is a single computer chip that combines multiple components—such as a processor, memory, graphics and communication controllers—into one compact package, like packing an entire desktop into a smartphone-sized tile. For investors, SoCs matter because they drive performance, power efficiency, and cost in consumer electronics and industrial devices; better SoC designs can give makers a competitive edge, influence product margins, and affect demand across supply chains.

robotaxis technical

"commercialization in robotaxis and its growing domain controller business."

Robotaxis are passenger vehicles that drive themselves using sensors and software, operating like a taxi without a human driver. For investors, they matter because they promise a new, potentially large revenue stream from automated ride-hailing and logistics while also carrying high costs, safety and regulatory risks, and heavy competition—similar to betting on a new public transit system that must prove it can run safely and profitably at scale.

domain controller technical

"announced its new generation autonomous driving domain controller, a high-performance"

A domain controller is a central computer server that stores and network’s user accounts, passwords and security rules and enforces who can access which systems and files—think of it as a building manager who checks IDs and opens doors. Investors should care because if this control point fails or is breached it can disrupt operations, expose sensitive data, harm customer trust, and create regulatory or recovery costs that affect a company's financial health.

tflops technical

"SoCs), achieving a combined maximum computing performance of 4000 FP4 TFLOPS."

TFLOPS (teraFLOPS) measures a computer or chip’s peak speed by counting how many trillion math calculations it can do each second, focusing on the kind of arithmetic used in scientific and AI workloads. For investors, TFLOPS is a quick way to compare raw compute power—like comparing how many cars a factory can produce per hour—and can signal a product’s competitiveness, potential market demand, and energy or cost trade‑offs.

AI-generated analysis. How Rhea-AI works. Not financial advice.

See more from StockTitan in Google Search and AI answers. Adds StockTitan as a preferred source · opens Google

Add on Google Not now

GUANGZHOU, China, April 25, 2026 /PRNewswire/ -- Pony AI Inc. ("Pony.ai") (NASDAQ: PONY; HKEX: 2026) today announced its new generation autonomous driving domain controller, a high-performance compute system designed for both Pony.ai's L4 autonomous driving platform and a broader set of customer applications across autonomous mobility. Developed in collaboration with NVIDIA, the new controller is built on the NVIDIA DRIVE Hyperion platform and powered by NVIDIA DRIVE AGX Thor with NVIDIA NVLink, supporting Pony.ai's next phase of commercialization in robotaxis and its growing domain controller business.

The new system is designed to deliver significant gains in AI computing performance, energy efficiency and support for the latest AI models, while meeting core L4 requirements such as multi-sensor fusion, full-scenario perception and high-complexity scenario understanding. It is also engineered to further enhance safety redundancy, system robustness and deployment flexibility.

Pony.ai expects the new platform to support a portfolio spanning multiple compute tiers and cooling solutions, enabling deployment across a wide range of autonomous applications. With flexible single-chip and multi-chip configurations, the platform is expected to incorporate NVIDIA NVLink to enable high-speed, low-latency communication between the two DRIVE Thor system-on-a-chips (SoCs), achieving a combined maximum computing performance of 4000 FP4 TFLOPS.

"Our collaboration with NVIDIA has supported several critical milestones in Pony.ai's autonomous driving journey," said Dr. James Peng, Founder and CEO of Pony.ai. "The next-generation domain controller built on NVIDIA DRIVE Hyperion will be a key enabler for the continued evolution of our L4 autonomous driving products and help accelerate large-scale commercialization."

"Autonomous driving systems are rapidly increasing in complexity, driving the need for scalable, high-performance compute platforms," said Rishi Dhall, Vice President of Automotive at NVIDIA. "Pony.ai has been a long-standing NVIDIA DRIVE customer and ecosystem partner, and we're pleased to see them build their next-generation domain controller on NVIDIA DRIVE Hyperion with DRIVE Thor accelerated compute to support advanced L4 autonomous driving applications."

The new platform builds on a long-standing collaboration between Pony.ai and NVIDIA that began in 2017. In 2022, Pony.ai launched its in-house developed automotive-grade computing unit powered by single or multiple NVIDIA DRIVE AGX Orin SoC, which was deployed in its sixth-generation Robotaxis. In 2025, the company began mass production of the world's first L4 Robotaxi domain controller equipped with four NVIDIA DRIVE AGX Orin SoCs. That controller now powers Pony.ai's seventh-generation (Gen-7) Robotaxis and provides a strong foundation for scalable, fully driverless operations.

As one of the few L4 autonomous driving companies globally with full-stack, in-house capabilities across both software and hardware, Pony.ai brings deep expertise in domain controller design. Its software-hardware co-design approach enables system-level optimization across performance, safety and cost.

Over the past year, Pony.ai has seen growing demand for its automotive-grade domain controllers across low-speed delivery, robosweeping, logistics, mining, autonomous shuttles and other robotics and intelligent mobility applications. Shipments of its "Fangzai" domain controller in 2025 surged by more than 500% year over year. Customers are spread across dozens of countries, including Germany, the United Kingdom, South Korea, Japan, and Switzerland.

Building on Gen-7's gains in cost efficiency, safety and reliability, Pony.ai has also achieved significant milestones in robotaxi commercialization, including unit-economics breakeven in two of China's major metropolitan markets. The company aims to expand its robotaxi fleet to more than 3,000 vehicles and its geographic footprint to more than 20 cities globally by the end of 2026.

About Pony AI Inc.

Pony AI Inc. is a global leader in achieving large-scale commercialization of autonomous mobility. Leveraging its vehicle-agnostic Virtual Driver technology, a full-stack autonomous driving technology that seamlessly integrates Pony.ai's proprietary software, hardware, and services, Pony.ai is developing a commercially viable and sustainable business model that enables the mass production and deployment of vehicles across transportation use cases. Founded in 2016, Pony.ai has expanded its presence across China, Europe, East Asia, the Middle East and other regions, ensuring widespread accessibility to its advanced technology.

Contact
Pony.ai: media@pony.ai

View original content:https://www.prnewswire.com/news-releases/ponyai-announces-new-generation-autonomous-driving-compute-platform-built-on-nvidia-drive-hyperion-302753517.html

SOURCE Pony AI Inc.

FAQ

AI-generated questions and answers. How Rhea-AI works. Not financial advice.

What is Pony.ai's new domain controller announced April 25, 2026 (PONY)?

It is a next‑generation L4 domain controller built on NVIDIA DRIVE Hyperion using DRIVE AGX Thor. According to Pony.ai, it supports single and multi‑chip NVLink configurations and a combined maximum of 4000 FP4 TFLOPS for advanced autonomy.

How much compute performance does Pony.ai claim for the new controller (PONY)?

Pony.ai cites a combined maximum computing performance of 4000 FP4 TFLOPS. According to Pony.ai, that figure is achieved via two DRIVE Thor SoCs connected with NVIDIA NVLink in multi‑chip setups.

What commercial markets will Pony.ai target with the new domain controller (PONY)?

The controller targets robotaxis and broader autonomous mobility applications including delivery, logistics, mining, and shuttles. According to Pony.ai, customers span dozens of countries across Europe and Asia.

How did Pony.ai's domain controller shipments perform in 2025 (PONY)?

Shipments of the Fangzai domain controller increased by more than 500% year‑over‑year in 2025. According to Pony.ai, demand grew across last‑mile delivery, robosweeping, logistics, and mining use cases.

What are Pony.ai's robotaxi expansion targets for 2026 (PONY)?

Pony.ai aims to expand its robotaxi fleet to more than 3,000 vehicles and enter over 20 cities globally by the end of 2026. According to Pony.ai, this builds on Gen‑7 commercialization progress and market breakeven in two cities.