Ambarella Expands Autonomous Driving AI Domain Controller Family With Two New SoCs

By Tiera Oliver

Associate Editor

Embedded Computing Design

January 10, 2024


SANTA CLARA, Calif.— Ambarella, Inc. announced during CES the latest additions to its CV3-AD family of automotive AI domain controllers—the CV3-AD635 and CV3-AD655 systems-on-chip (SoCs).

The new CV3-AD635 and CV3-AD655 SoCs are fabricated in Samsung’s advanced 5nm automotive process technology. TheCV3-AD635 is designed to support a sensing suite that includes multiple cameras and radars to enable mainstream L2+ feature sets, such as highway autopilot and automated parking, in addition to meeting the GSR2 and NCAP standards. The CV3-AD655 is designed to enable advanced L2+ (also called L2++) with urban autopilot, and support for additional cameras, radars, and other sensors.

With the previously announced flagship CV3-AD685 SoC, which targets L3/L4 systems—along with the China-focused CV72AQ SoC—the CV3-AD family now covers the full range of AD and ADAS solutions, from mainstream to premium passenger vehicles. Per the company, this solution offers a wide range software-compatible AI domain control family, spanning an 18x AI performance increase from the CV72AQ to the CV3-AD685.

All SoCs in the CV3-AD family feature Ambarella’s CVflow 3.0 AI accelerator. This proprietary AI engine combines neural network acceleration—including support for the latest NN architectures such as transformers and bird’s-eye-view (BEV) networks—along with general vector processing capabilities for computer vision and radar processing.

The CVflow’s AI processing performance and power efficiency enable CV3-AD customers to design vehicles with feature sets based on AD and ADAS capabilities, while reducing thermal-management cost and complexity. The low power consumption also allows customers to deploy electric vehicles with small batteries, reducing cost and weight; or to alternatively increase the EV range using the same battery.

“The strategic partnership between Continental and Ambarella is bringing full-stack vehicle system solutions to the road—beginning with 2027 SOPs—that combine maximum performance and industry-leading energy efficiency,” said Ismail Dagli, Head of the Autonomous Mobility Business Area at Continental. “Based on Ambarella’s CV3-AD family, our joint solutions enable safer mobility thanks to holistic environmental perception; making them particularly well suited for the growing amount of sensor data in software-defined vehicles. Together, we are shaping the path toward Vision Zero and autonomous mobility.”

Ambarella’s CV3-AD SoC portfolio meets the processing requirements for automotive OEMs’ catalogs, from mainstream to premium models, by scaling the performance of each members’ compute engines. According to the company, the CV3-AD655 offers double, and CV3-AD685 six times, the NN processing performance of the CV3-AD635. The CV3-AD635 integrates four Arm Cortex A78AE CPU cores, and the CV3-AD655 eight cores. A dual-core, lockstep pair of Cortex-R52 CPUs is also included in both the CV3-AD635 and CV3-AD655. These SoCs target ASIL-B at the chip level, via an ASIL-D safety island. Also integrated on-chip is a GPU to render 3D surround views.

Portability across the CV3-AD family is achieved by sharing the same architectural approach, SDK, and tools. Ambarella’s full software stack for autonomous and semi-autonomous driving is optimized for all CV3-AD family members, as the stack and SoCs were designed in conjunction.The CV3-AD family’s on-chip image signal processor provides imaging in low-light conditions, including high-dynamic-range processing to extract maximum image detail in high-contrast scenes. All CV3-AD family members also provide high-resolution video recording and streaming at low bit rates with H.264 video encoding.

The integrated hardware security module provides isolation of different domains and secure software provisioning, and a suite of cybersecurity features, including: asymmetric/symmetric crypto acceleration, secure storage and key provisioning, encrypted CVflow AI acceleration tasks, true random number generator, one-time programmable memory, DRAM scrambling, and DRAM virtualization.

CV3-AD635 and CV3-AD655 engineering samples are expected to be available in Q1 2024. At Ambarella’s invitation-only exhibition during CES in Las Vegas this week, the company will provide fully autonomous test drives in its R&D vehicles to demonstrate the CV3-AD family’s capabilities

For more information, please visit

Tiera Oliver, Associate Editor for Embedded Computing Design, is responsible for web content edits, product news, and constructing stories. She also assists with newsletter updates as well as contributing and editing content for ECD podcasts and the ECD YouTube channel. Before working at ECD, Tiera graduated from Northern Arizona University where she received her B.S. in journalism and political science and worked as a news reporter for the university’s student led newspaper, The Lumberjack.

More from Tiera