Chip designer Arm is bringing its Nuverse architecture to the automotive sector with the first Armv9 processors for use in vehicles, claiming they will deliver better performance for AI.
The chip designer introduced Automotive Enhanced (AE) versions of two Cortex-A processors, as well as a Neoverse V3AE high-performance core, a real-time processor and an image signal processor, all designed for automotive use cases.
As usual, these are designs for Arm’s chip manufacturing customers to license for their silicon products, and so are unlikely to filter into production vehicles for many years.
The automotive market, according to Arm, is going through a period of great change, with vehicle electronics and software becoming increasingly complex.
To try to combat this, the chip designer said it is bringing server-class Nuverse Tech to the market, along with the AI, security and virtualization features of Armv9-based Cortex-A products.
Neoverse V3AE is aimed at use cases like autonomous driving and high-end driver assistance systems that demand high single-threaded performance, so it’s fortunate that it’s based on the V3 core announced last month, billed as was then granted by Arm. “Highest single-threaded performance Evercore ever.”
This means the Core has the same core CPU design and performance as Arm’s latest cloud and data center designs, to meet the growing demand for computing in vehicles, and features Automotive Enhanced (AE). Adds
Similarly, the Cortex-A720AE and Cortex-A520AE are based on the Cortex-A720 and Cortex-A520 that the chip designer introduced last year as part of its TCS23 system-on-chip (SoC) platform, but with additional AE features. With.
These AE features include transient fault protection (TFP) and split-lock support, allowing a pair of cores to be configured to operate independently as normal (split mode), or locked for fault-tolerant operation. Step is done.
The Cortex-A720AE targets driver assistance and “digital cockpit” systems, while the Cortex-A520AE is intended for use cases where workloads need to be processed as efficiently as possible.
It says that Arm has seen a 30 percent performance increase when moving automotive workloads from the older Cortex-A78AE to the Cortex-A720AE.
As with the TCS23 platform, the Cortex-A720AE and Cortex-A520AE can be combined in a cluster configuration of up to 14 cores thanks to the DSU-120AE. It is a version of Arm’s DynamIQ shared-unit glow logic that now has the AE features required for automotive systems.
Also introduced is the Cortex-R82AE, a real-time core designed for functional safety applications, which may include monitoring other processors or controlling vehicle functions. Arm says the Cortex-R82AE brings 64-bit processing to the line for the first time.
The final core is the Mali-C720AE Image Signal Processor (ISP), designed for in-vehicle vision systems. Compared to previous designs, it has multiple parallel pipelines that can be tuned differently to support human vision or computer vision use cases. Arm claims that this reduces the time it takes to process images, speeding up the overall reaction time of the compute system.
The chip designer has expanded its portfolio with AE-enabled versions of components such as network-on-chip (NoC) interconnects and coherent mesh networks.
Arm also says it’s introducing virtual prototyping, starting with its latest Arm AE automotive products. The company claims this will allow software developers to start their work before physical silicon is available, reducing time to market by up to two years.
“The expansion of the Arm AEIP portfolio will power and enhance the next generation of vehicles,” wrote Tom Conway, Arm’s senior director of product management for automotive, on the company’s blog.
Conway said the company is building tech to respond to changes in the automotive industry, including AI-powered workloads and increased security and safety features. ®