Samsung to produce 5nm chipsets for self-driving vehicles

Samsung recently announced its entry into the autonomous vehicle industry. However, the company will not be producing its self-driving vehicles like Xiaomi. Instead, Samsung plans to supply chips that AI will use to control self-driving systems. Ambarella, a Tier-1 automotive supplier based in the US, will be Samsung's customer. Together, the two companies promise to bring revolutionary changes to the next generation of autonomous vehicles.

Ambarella is a company that started in 2004 developing video encoders for professional broadcast services but later transitioned to developing low-power video compression chips for consumer use. Their technology has been used in products like GoPro cameras, Nest Dropcams, Garmin dashcams, and DJI Phantom drones.

In recent years, Ambarella has acquired companies from the automotive field, incorporating their technology into solutions for autonomous vehicles. They have partnered with Incepto Technology to provide an automotive-grade central computing platform and with Continental and Samsung to focus on AI-based software and hardware systems for advanced ADAS and fully automated driving.

According to a blog post, Samsung will manufacture a 5nm chip for Ambarella, which will be the first one based on the CV3-AD family of central controllers. The chip comes with an ARM Cortex A78AE, R52 CPU cores, an automotive-grade GPU, a dedicated security module, and an advanced ISP for processing multiple camera inputs. Its algorithm-first architecture supports a complete software stack for Level 2+ to Level 4 autonomous driving.

Ambarella's CV3-AD685 System-on-Chip (SoC) is used for advanced driver assistance systems, which serve as the brain of an autonomous and self-driving vehicles, Samsung said.

"Ambarella and Samsung Foundry have a rich history of collaboration, and we are excited to bring their world-class 5nm technology to our new CV3-AD685 SoCs," said Fermi Wang, President and CEO at Ambarella. "Samsung’s proven automotive process technology allows us to bring new levels of AI acceleration, systems integration and power efficiency to ADAS and L2+ through L4 autonomous vehicles."

Ambarella selected Samsung's 5nm process because of its optimisation for automotive-grade semiconductors. Apparently, the new chip can handle neural network processing for 4D imaging radars, computer vision, and deep sensor fusion combined with path planning in ADAS.

"Samsung brings 5nm EUV FinFET technology to automotive applications for unprecedented ADAS and vision processor performance," said Sang-Pil Sim, Executive Vice President and Head of Foundry Corporate Planning at Samsung Electronics. "With Tier-1 automotive suppliers already adopting the technology, we believe other automotive companies will also consider using the Ambarella CV3-AD SoC product family manufactured in Samsung’s 5nm process."