The Automaker's Glovebox Manual

The Automaker's Glovebox Manual.

The handbook for a disrupted automotive industry.

Auto Hero ContentContents

Scalability: Adopting a scalable compute platform for AI-defined vehicles.

Creating a scalable platform for passenger car manufacturing means addressing the challenges of transitioning from traditional fragmented electronics architectures to scalable platforms that support AI-enabled software-defined vehicles (SDVs). Vehicle manufacturers must move toward a centralized compute hardware platform that can interface with standardized software, allowing for a single brand differentiating software interface.

This approach requires taking a view on components as a full solution, from hardware to application, which in turn enables:

More ADAS features to be integrated in vehicles.
Monetization opportunities for software products.
An improved driving experience.

Moreover, advanced driver-assistance features are growing in number. In the U.S. alone, it is estimated that these technologies have the potential to prevent around 21,000 fatalities per year. At the same time, these features are highly complex, and require a more diverse set of technologies and higher level of compute power, and largely fall within ISO26262 ASIL-B or ASIL-D safety standards.

The Arm Solution: Automotive Enhanced Technology

Arm automotive technologies feature cutting-edge solutions for ADAS and autonomous driving. In March 2024, Arm and its partners introduced the Arm Automotive Enhanced (AE) processors and new virtual platforms. These innovations are critical for not only delivering ADAS solutions more efficiently and quickly, but also for offering exceptional performance scalability, power efficiency, and functional safety features that ease the path to ASIL B(D) and ASIL D ISO 26262 certification. The newest additions to the AE suite include:

Arm Neoverse V3AE CPU: A server-class processor with top-tier single-thread performance.
Armv9 Cortex-A CPUs: Two advanced processors designed for performance (Arm Cortex-A720AE CPU) and efficiency (Cortex-A520AE CPU).
Arm Cortex-R82AE CPU: Arm’s first 64-bit real-time CPU, providing high performance and determinism.
Arm Mali-C720AE: An image signal processing (ISP) core, optimized for AI-driven computer vision.

Together, these products provide an end-to-end computing foundation for silicon vendors, vehicle manufacturers, and other partners across the automotive supply chain.

Learn how Arm helped accelerate time to market, cut latency, and optimize CPU use for LeddarTech.

As SDVs naturally provide the foundation for AI-defined vehicles, this scalable compute foundation becomes even more critical. ADAS features are no longer reactive—they are becoming predictive. AI-defined vehicles will use real-time inference from radar, lidar, and camera data to anticipate driver behavior and road conditions before taking corrective action. This requires high-performance, low-latency compute at the edge. Arm’s AE technologies provide the processing power and safety features needed to enable the next generation of AI-based perception and decision-making in a scalable, power-efficient architecture.

Latest Arm Technology Advances

As we look to the future, the complexity of systems will continue to increase with more compute, more AI features, and safety becoming even more critical. The Arm Zena Compute Subsystem (CSS) offers pre-integrated and validated configurations of Arm AE IP. These configurations are optimized for performance, power, and area using advanced foundry processes. By integrating all the necessary components, Arm Zena CSS will help our partners bring products to market faster than ever.

This includes the Arm Neoverse V3AE CPU, which delivers server-class performance for automotive applications, enabling real-time AI processing that is paramount for complex decision-making in ADAS and autonomous driving.

Meanwhile, Armv9 Cortex-AE processors support high-performance computing and a range of AI workloads, allowing for faster and more accurate object detection, path planning, and driver monitoring systems.

Additionally, chiplets offer numerous intriguing SoC design possibilities for the ecosystem. Arm, through our partner ecosystem, delivers a versatile compute platform to support the expanding chiplets ecosystem. Our solution stands out by offering multiple reusable IP components that can be integrated into larger systems by partners, as showcased across our new Automotive Enhanced IP technologies.

Discover how Arm is powering autonomous driving solutions.

ADAS & ISO 26262: Best practices for integration, testing, and validation.

Changing consumer demands for safer, smarter, and more connected vehicles has resulted in an unprecedented evolution of the automotive industry. This is coupled with the rise of autonomous driving, electric vehicles, and ADAS. As a result, the overall complexity of automotive systems has increased significantly, bringing new safety challenges that must be addressed to protect consumers, while helping to ensure an optimal driving experience.

Streamlining ADAS software integration, testing, and validation involves addressing several key challenges. One of the primary challenges is the complexity of integrating ADAS features into existing vehicle architectures, which requires sophisticated software coordination. Additionally, ensuring real-time processing for safety-critical functions is crucial, as any delays can compromise the system's reliability.

Our approach to safety and performance delivers:

Enhanced performance and efficiency: High-performance processors from the Arm Automotive Enhanced (AE) IP portfolio.
Seamless integration: A comprehensive support package, including software test libraries (STLs), tools, and compilers.
Accelerated Development: Virtual platforms for pre-silicon software development and testing, helping to reduce time-to-market.
Compliance with safety standards: Adherence to UNECE GSR, ISO 26262, ISO 21448 (SOTIF), and ISO 21434 standards across our Safety Ready solutions and AE processors.
Robust ecosystem: SOAFEE (Scalable Open Architecture for Embedded Edge) is an initiative to help reduce software complexity and enable a multivehicle platform, single software approach.

Safety, Performance, and Efficiency in an SDV Era

The unprecedented advances in software and AI that are defining software-defined vehicles (SDVs) require new levels of performance, efficiency, safety and security. This is a challenge that we set out to meet through a brand-new suite of leading-edge processors that expands the portfolio of Arm Automotive Enhanced (AE) IP.

The intended product based on the target solution must comply with relevant regulations and safety standards. For example, an ADAS solution must adhere to UNECE GSR and standards like ISO 26262, ISO 21448 (SOTIF), and ISO 21434, to help ensure correlation between safety and security.

Learn more about the top considerations for automotive safety.

Software-First Approach

The availability of development environments for our latest AE IP means that software can be developed and tested before the physical silicon is ready, hleping to cut development time and leading to a faster time-to-market.

A prime example is LeddarTech, which optimized its ADAS perception and fusion algorithms for the next-generation Arm AE IP, specifically the Armv9-based Cortex-A720AE CPU. By using pre-silicon platforms and Armv9 software readiness, it leveraged new features like SVE2 (Scalable Vector Extension 2), achieving more than a 30% boost in performance compared to the previous Cortex-A78AE CPU.

Meanwhile, Arm, alongside other industry leaders such as AWS, Continental, BOSCH, and Red Hat, founded SOAFEE for Embedded Edge in 2021, which is driving automotive industry standardization to help reduce software complexity. Since then, a vast ecosystem has grown and is using SOAFEE methodologies to implement a multivehicle platform, single software approach.

120 leading automotive companies, including Geely, GM, and Tata Motors are revolutionizing the automotive industry with SOAFEE. Learn more.

Arm, in collaboration with Codethink, has tackled a crucial part of this integration puzzle by incorporating Arm Software Test Libraries (STL) into live Arm Cortex-A application processor software stacks. This specifically addresses the integration at Exceptional Level 3 (EL3), within Trusted Firmware-A (TF-A) for runtime utilization, which allows developers to enhance the integrity and reliability of their systems.

Read more here

The emergence of AI-defined vehicles will add a new layer of complexity to ADAS validation. Features like predictive collision avoidance, adaptive cruise control, and driver behavior monitoring rely on real-time AI inference under strict safety constraints.

These applications must be tested across a wide range of dynamic environments before deployment. With virtual platforms and full-stack software solutions from across Arm’s ecosystem, developers can simulate and validate AI-powered ADAS stacks before silicon is available, ensuring performance and compliance with standards like ISO 26262 and ISO 21434, while reducing time-to-market by up to two years.

Efficiency: Balancing high-performance hardware with cost in SDVs.

The increase in digitalized safety-critical vehicle features and change in vehicle OEM business models generates a high degree of complexity and underscores the need for a software-defined approach. However, this is leading to a need for high-performance hardware that can be costly to implement. For OEMs, these high-performance computing systems must be cost effective for mass production, while maintaining performance.

To better manage the multiple compute elements and increasing software complexity, which includes support for over-the-air updates, there is a push toward revamping the vehicle architecture. Traditionally, discrete electronic control units (ECUs) manage specific functions on the vehicle. These ECUs are then added to the architecture to support new features and functions. However, this model is neither scalable nor suitable for SDVs, as it substantially increases the challenges of managing complex software and software updates.

As a result, there is a push toward a more centralized architecture. This means having discrete and spatially located ECUs managing multiple functions that are consolidated into fewer powerful zonal controllers that run multiple software workloads.

This shift allows OEMs to protect software investments and enable post-production monetization across different car segments. Additionally, the evolution of the automotive EE-architecture to a zonal architecture with central compute supports the trend toward further software integration in cars. This minimizes the overall number of ECUs in the car, which reflects the requirements of the AI-enabled SDV for ongoing software updates and maintenance.

SOAFEE

The functionalities of SDVs are driven by software features and AI enhancements, advancing everything from the user experience to autonomous driving. While this allows the industry to reimagine vehicles as safer, more efficient, and more adaptable than ever before, the complexity challenge also presents.

Learn more about zonal and centralized compute architecture.

The good news is that we know how to get there and SOAFEE plays a key role in this journey. SOAFEE provides a common device virtualization framework that decouples software implementation from diverse hardware targets. This allows for greater flexibility and scalability across different vehicle variants and generations, helping to reduce the dependency on specific hardware.

Arm Neoverse V3AE

Arm Neoverse V3AE, the first server-class processor for automotive markets, plays a critical role in addressing the shift from a hardware-focused to a SDV business model by offering advanced computing solutions that balance high performance with cost constraints. One of the key benefits of Arm Neoverse V3AE is its ability to standardize interfaces between software stacks and underlying hardware. This standardization simplifies the integration of different vendor solutions, helping to reduce the overall costs associated with developing and maintaining multiple hardware-specific software versions.

Additionally, the platform’s powerful and scalable computing capabilities support advanced automotive applications, such as ADAS and autonomous driving. By using the high-performance capabilities of Arm Neoverse V3AE, vehicle manufacturers can help ensure that their vehicles are equipped with the necessary computational resources to support sophisticated software features, while keeping hardware costs manageable.

Visit developer.arm.com for more details on Arm Neoverse V3AE.

A New Approach to Building Silicon

Adopting a modular and scalable architecture for ADAS platforms can enhance their adaptability. The Arm Zena Compute Subsystem (CSS) provides a faster path to building chiplet-based or monolithic silicon designs for ADAS.

Its pre-integrated platform reduces time to silicon by up to 12 months, supports both real-time and AI-defined workloads, and includes safety-certified components aligned to ISO 26262 and ISO/SAE 21434. This modular approach allows OEMs to customize and scale their ADAS solutions to meet specific regional requirements and evolving global regulations.

Additionally, chiplets offer numerous intriguing SoC design possibilities for the ecosystem. Through our partner ecosystem, Arm delivers a versatile compute platform to support the expanding chiplets ecosystem. Our solution stands out by offering multiple reusable IP components that can be integrated into larger systems by partners, as showcased across our new Automotive Enhanced IP technologies.

As AI-defined vehicles introduce more advanced ADAS workloads, from object anticipation to driver attention tracking, the need for real-time, high-throughput processing grows. However, delivering this intelligence at scale requires compute platforms that are not only powerful, but also cost-effective. Arm Zena CSS enable silicon vendors and OEMs to scale AI capability across car tiers without a full redesign, balancing AI performance with commercial viability.

System76 and Ampere have introduced Thelio Astra, an Arm64 developer desktop designed to revolutionize the Arm Linux development process for automotive applications. Learn More.

Learn why Arm Neoverse V3AE is the CPU foundation for the automotive industry.

Real-time & Secure ADAS: Optimizing processing and cybersecurity within power limits.

With the advent of AI-enabled SDVs, which are expected to contain billions of lines of code and significantly enhanced connectivity, the automotive attack surface is expanding and evolving. This growth has profound implications for automotive cybersecurity, as new vulnerabilities are continuously reported to the MITRE Common Vulnerabilities and Exposures (CVE) database, with the number increasing annually. To mitigate the risks posed by these security flaws, the automotive industry is proactively building comprehensive defence-in-depth strategies across the entire automotive ecosystem.

The integration of ADAS has increased the amount of data and value of assets inside the vehicle. These include sensor and actuator data, AI models and algorithms used for perception and object classification, graphic-intensive computing like 360-degree cameras, and various mixed-critical considerations. The security threats are then amplified further due to the increasing attack surface from the large amount of software in AI-enabled SDVs and high-potential impact of an attack, since both ADAS and AD directly influence vehicle control.

Protecting the Digital Cockpit

At the same time, other applications within the SDV, like the digital cockpit and IVI, are becoming increasingly integrated with ADAS, which is making the creation and subsequent management of these systems more complex. Moreover, due to the increasing connectivity features and demands of these systems, attack surfaces are becoming larger.

The digital cockpit has the largest attack surface as it covers multiple vectors, including cloud connectivity, connectivity to personal devices like smartphones, USB plug-ins, and the ability to download applications. There are also various motivations to hack the digital cockpit as it contains personal data that could be valuable to hackers, like payment details. For IVI, the main security risk is offering a gateway to the rest of the vehicle for theft or control of the vehicle. This provides a route for ransom or denial of service attacks.

The Arm Automotive Enhanced (AE) IP Portfolio is designed to meet the heightened demands for performance, safety, and security in AI-enabled SDVs. These processors address several critical security challenges in automotive applications, including:

Increasing software complexity.
A highly diverse software supply chain.
Feature enablement hacking.
Ransomware.
Securing high-speed communication.
Managing privacy for passengers and across different environments.

State-of-the-Art Security Practices

The latest Arm AE processors adopt key Armv9 defensive execution technologies and architecture features that protect against the consequences of vulnerable or malicious software. Pointer Authentication (PAC), Branch Target Identification (BTI), and Memory Tagging Extension (MTE) help to overcome the risk exposed by growing lines of code by protecting the integrity of the software control flow and reducing the impact of memory safety bugs. This is important for automotive markets because there are still vast amounts of legacy code written in memory-unsafe languages, like C, which can be ported into future SDVs.

In AI-defined vehicles, real-time ADAS decision-making becomes both a performance and security imperative. AI models powering features like lane merging prediction, drowsy driver alerts, or adaptive cruise control must operate in milliseconds while remaining isolated from other critical systems. Arm’s architecture supports this dual requirement through compute-efficient inference at the edge and hardware-enforced trust boundaries via technologies like RME and TrustZone. This allows safety-critical ADAS AI to be deployed confidently, without opening new attack surfaces or compromising system integrity.

Additionally, Arm follows state-of-the-art product security practices and standards, like ISO/SAE 21434, to ensure that security risks are managed during the idealization, development, and post-development of all products. Arm provides a set of supporting security materials for automotive partners to simplify the integration of our off-the-shelf components into ISO/SAE 21434-compliant designs.

Built into the Arm architecture are scalable isolation technologies that segregate diverse workloads with minimal performance impact. A trend in the automotive industry is to run mutually distrusted software components from varied sources of origin on the same computing platform. Isolation technologies help support this aim by offering a strong enforcement of well-defined trust boundaries. Examples of these technologies include Arm TrustZone, S-EL2 and Realm Management Extension.

Arm co-founded and continues to contribute toward standard security APIs, including the PSA Certified Crypto API, which works as a contract between firmware developers and hardware vendors. This allows developers to focus on designing firmware rather than having to understand proprietary hardware rules with each new integration. Meanwhile, for hardware vendors, standard APIs are a way to remove barriers to entry, helping them to instead focus on valuable commercial differentiation.

Learn how Arm is helping secure the automotive industry.

IVI Integration: Balancing constraints, durability, and performance in an AI-defined vehicle.

IVI systems are becoming increasingly sophisticated, integrating entertainment, navigation, connectivity, and driver assistance features to create a smarter, more interactive experience for drivers and passengers.

Recent trends include the adoption of larger and more complex screens, AI-powered voice recognition, and driver recognition technologies that help enhance safety and convenience.

Additionally, there is a growing demand for seamless connectivity between vehicles and personal devices, driven by consumer expectations for a connected car experience.

The Arm Automotive Enhanced (AE) Portfolio offers a comprehensive suite of solutions that effectively help address the challenges of integrating IVI systems within a vehicle’s design constraints.

Seamless Integration of IVI Systems

The latest additions to the Arm AE portfolio include high-performance Armv9 AE CPUs, which are optimized for automotive applications. These processors deliver the necessary computational power for advanced IVI systems, while maintaining low power consumption, which is crucial for managing the thermal constraints and battery life in vehicles.

Additionally, the portfolio’s scalability allows for the development of a wide range of systems-on-chip (SoCs) tailored to specific automotive applications and components, including digital cockpits and ADASs. This flexibility ensures that IVIs can be seamlessly integrated into various systems in the vehicle.

Alongside CPUs, Arm AE IP also includes advanced interconnects, memory management units (MMUs), and image signal processing cores, which are essential for handling the complex data flows and multimedia requirements of modern IVI systems. This helps ensure seamless operation and high-quality user experiences.

Ensuring long-term durability

Maintaining the durability of IVI systems involves comprehensive durability testing, which assesses the system’s ability to withstand various operational stresses over time. This includes vibration testing, thermal cycling, and humidity testing to ensure that all components can endure the harsh conditions they might face in a vehicle.

Here, adhering to industry reliability standards is crucial. Building IVI systems on the Arm AE IP portfolio, which adheres to rigorous testing and reliability standards, automotive manufacturers can ensure long-term durability and performance.

To accelerate deployment of next-gen IVI systems, the Arm Zena Compute Subsystem (CSS) provides a complete, software-ready compute platform that includes optional GPU and ISP integration, real-time and application processing support, and pre-integrated safety and security features. Combined with pre-integrated firmware and full stack software support, Arm Zena CSS enables faster time-to-market and easier feature deployment across vehicle lines.

Flexible, Scalable Software

Meanwhile, SOAFEE for Embedded Edge provides a robust framework to address the challenges of seamlessly integrating IVI systems. The SOAFEE architecture supports the development of SDVs through decoupling hardware and software, which allows for a more flexible and scalable integration of IVI systems.

This decoupling is achieved through virtualization technologies, which abstract vehicle hardware components like ECUs, sensors, and actuators, and enable software to control these resources more efficiently.

The SOAFEE cloud-native approach helps in the development of IVI applications in a virtual environment before they are validated in real-world conditions. This not only helps reduce development time, but also helps ensure that applications can be seamlessly integrated into various vehicle architectures.

Additionally, SOAFEE support for mixed-criticality workloads helps ensure that both safety-critical and non-safety-critical applications can coexist and operate reliability within the same system.

AI-defined vehicles are taking IVI to the next level, transforming it from a control hub into an intelligent, responsive in-cabin assistant. Future systems will feature conversational interfaces, adaptive content, and emotion-aware interactions powered by AI at the edge. These features require high-performance, low-power compute platforms that can process and personalize data in real time.

Arm’s Automotive Enhanced (AE) technologies, with support for heterogeneous workloads across CPU, GPU, and accelerators, delivers the performance, scalability, and energy efficiency needed to support these advanced IVI experiences—without compromising thermal or power constraints.

Learn more about SOAFEE through GitLab.

Amazon Web Services (AWS) on Arm

AWS uses Arm-based Graviton processors to deliver high-performance, cost-effective, and energy-efficient cloud computing solutions. These processors are optimized for automotive applications, offering the necessary computational power for advanced IVI systems, while maintaining low power consumption.

Additionally, AWS supports the use of the Android automotive operating system (OS) on its EC2 instances, allowing vehicle manufacturers and tier-1 suppliers to develop and test their infotainment applications at scale.

A New Way to Build Silicon

Adopting a modular and scalable architecture for ADAS platforms can help enhance their adaptability. The Arm Zena Compute Subsystems (CSS) offers an accelerated path to building chiplet-based designs through enhanced computing and integration capabilities. This modular approach allows OEMs to customize and scale ADAS solutions to meet specific regional requirements and evolving global regulations.

Our solution stands out by offering multiple reusable IP components that can be integrated into larger systems by partners, as showcased across our new Automotive Enhanced IP technologies.

Click here to learn more about in-vehicle infotainment from Arm.

IVI Systems: Best strategies to balancing computational power and security.

The demand for modern IVI systems to have high-definition displays, AI-powered features, and seamless user experiences is driving the need for powerful and efficient computing solutions. Additionally, as IVI systems become more connected, the importance of robust cybersecurity measures to protect personal and financial data is increasing.

Therefore, embedding security into every stage of the software development lifecycle (SDLC) is crucial, including threat modelling, secure coding practices, and regular security testing. Moreover, helping to ensure that over-the-air (OTA) updates are secure is also essential for maintaining the security of IVI systems throughout their lifecycles. This involves using secure communication channels and verifying the integrity of updates before applying them.

By leveraging Arm technology, OEMs can develop IVI systems that meet the high computational demands of modern applications, while maintaining stringent security standards to deliver a safer, more enjoyable driving experience.

Arm’s approach to automotive security includes defensive execution technologies, such as Pointer Authentication and Branch Target Identification (PACBTI), which protect against control-flow attacks. These technologies help ensure that the execution of code is secure and resistant to tampering.

MTE is an extension to the Arm architecture to help detect errors in software use of memory, such as out of bounds accesses and those tricky to find use of the three errors. In the video above, Arm experts Graeme Barnes and Travis Walton describe how MTE works.

Arm also uses MTE (Memory Tagging Extension) and RME (Realm Management Extensions) to isolate different software components, which helps prevent malicious code from affecting critical system functions. MTE is an extension to the Arm architecture and helps detect errors in software use of memory, such as out of bounds accesses.

Meanwhile, Arm's suite of platform security services, including secure boot and run time security, help ensure that the system starts and operates securely. Developers can also use standard security APIs to help ensure interoperability and ease of integration for a consistent security framework across different automotive platforms.

Finally, Arm’s built-in security solutions, including Arm TrustZone, provide robust protection against cyber threats, helping to ensure the integrity and reliability of IVI systems.

As IVI systems evolve into intelligent AI-defined interfaces, safeguarding personal data becomes even more critical. Features like driver recognition, behavior-based personalization, purchasing, and voice-based control generate and process sensitive data that must be protected in real time.

Arm’s safety-ready AE platforms are built with security-first architecture, supporting ISO/SAE 21434 compliance and integrated defences such as TrustZone, MTE, and RME. These features ensure that advanced IVI capabilities remain protected across their lifecycle, from development to OTA updates.

Learn how Arm supports and secure AI-enabled SDV.

Functional Safety

The intended product must comply with relevant regulations and safety standards. For example, an ADAS solution should adhere to UNECE GSR and standards like ISO 26262, ISO 21448 (SOTIF), and ISO 21434, the help ensure the correlation between safety and security.

Arm Safety Ready Portfolio

Arm Safety Ready solutions support a wide range of technologies developed by our world-leading partner ecosystem for the automotive industry. Arm delivers safety-certified IP, software test libraries, software development kits, tools, operating systems, and compilers to aid the seamless integration of safety solutions into the customer-specific design.

The comprehensive Arm Safety Ready portfolio is developed through the application of a well-defined process, combined with safety tools and software components for faster than ever safety development.

Arm provides best-in-class documentation and certifies our product portfolio with independent assessment teams as part of the Arm Safety Ready strategy.

Learn why Arm is the right architecture choice for IVI systems.

From an Arm AE product perspective, meeting ISO 26262 standards is crucial and vehicle lifecycle safety must underpin partner solutions. Arm Safety Ready solutions are designed with a “safety first” approach, incorporating advanced safety mechanisms and comprehensive packages to help partners achieve ISO 26262 compliance.

IVI Security: Protecting user data in multimedia, navigation, and more.

Handling Multimedia, Navigation, and Voice Recognition

Arm technologies provide a comprehensive framework that tackle the challenge of designing IVI systems to handle multiple features simultaneously.

Voice Recognition

Arm AI-driven speech recognition technologies, such as those implemented on Arm Cortex-M series microcontrollers (MCUs), enable efficient and accurate voice recognition. These systems can operate offline, helping to ensure user privacy and reduce latency. The use of deep neural networks allows for high accuracy and robustness in various environments, making it suitable for in-vehicle applications.

In AI-defined vehicles, IVI systems will no longer just respond but anticipate. From emotion-sensitive playlists to gaze tracking and fatigue detection, these systems will use edge AI to adapt to both driver and passenger states in real time. This introduces new demands on privacy, inference security, and workload partitioning. The Arm compute platform and security solutions help isolate these AI applications, ensuring that multimedia, navigation, and personal data remain securely contained, even as intelligence, capabilities and use cases grow.

Heads-Up Display (HUD)

Meanwhile, Arm Mali GPUs are designed to deliver high-performance graphics with low-power consumption, which makes them ideal for HUDs in vehicles. These GPUs support advanced rendering techniques that can handle complex visual data, helping to ensure the HUD provides clear and responsive information to the driver.

Current and future in-vehicle driver and passenger experiences are being greatly enhanced through AI-enabled human machine interfaces (HMI) technologies.

Hear experts from Arm and Panasonic Automotive Systems on the tech driving this future.

Navigation

Arm Automotive Enhanced (AE) processors, such as those in the Arm Cortex-A series, provide the computational power needed for real-time navigation and mapping applications. Mapbox, a leading platform for powering location experiences, has developed its virtual head unit (VHU) with Arm and Corellium. This creates virtual prototypes of the Arm-based in-vehicle hardware before seamlessly integrating these with Mapbox’s navigation stack. Automotive OEMs can use the new VHU to build maps however they want, and then test and render it quickly before deployment.

Ensure User Data Protection with Arm

Arm uses MTE and RME to isolate different software components, which in turn helps prevent malicious code from affecting critical system functions.

These technologies create secure compartments within the system, isolating different software components. The IVI system can be imagined as a high-security building with multiple rooms, each with its own security measures. Even if one room is compromised, the others remain secure, protecting sensitive information from malicious attacks.

Arm also promotes the use of standard security APIs, which help ensure interoperability and ease of integration across different automotive platforms. This is the same as having a universal security protocol that all devices in the car adhere to, creating a consistent and robust security framework. It helps ensure that no matter what brand or model of car is driven, the data is protected by the same high standards.

In addition to these measures, Arm integrates real-time intrusion detection systems (IDS) into the vehicle's ECUs. These systems act like a sophisticated alarm system, detecting any suspicious activities or attempts to compromise the IVI system in real time. If an intruder tries to break in, the IDS immediately raises an alert, allowing for swift action to prevent any data breaches.

Learn why Arm is the foundation for the evolving in-vehicle experience.

Vehicle Computing: Maintain performance and reliability when integrating vehicle electronics.

Modern vehicles are increasingly integrating complex computing systems such as powertrain, IVI, and ADAS to enhance functionality and user experience. This shift toward centralized compute architectures that execute multiple applications is enabling more efficient and scalable vehicle designs and also facilitating faster decision making.

One effective strategy is transitioning from distributed ECUs to a centralized electrical/electronic (EE) architecture. Arm high-performance solutions that feature in SoCs are designed to support such a centralized architecture, helping to reduce latency and improve data processing efficiency. Moreover, Arm solutions enable the integration of various applications, such as ADAS, IVI, and powertrains, on independent hardware boards. This shift allows OEMs to protect software investments and realize post-production monetization across different car segments.

Additionally, the evolution of the automotive EE-architecture to a zonal architecture with central compute supports further software integration in cars. This helps minimize the overall number of ECUs in the car, which reflects the requirements of the AI-enabled SDV that need ongoing software updates and maintenance.

Learn more about zonal and centralized compute architecture.

Meanwhile, using high performance SoCs designed for automotive applications can also support the integration of multiple functions. Arm’s Automotive Enhanced (AE) processors deliver AI-accelerated compute capabilities, helping to ensure that both safety-critical and non-critical applications run smoothly. These processors are built to handle mixed criticality workloads and provide the necessary performance and reliability for complex automotive systems.

Arm-based virtual platforms have been developed for ADAS and autonomous driving by our ecosystem partners to enable both accelerated product development and reduced time to market. With these platforms, designers can start work immediately, rather than having to wait for physical hardware to become available before adding a feature to a prototype. The ADAS stack can be used ‘as is’ from IP to Arm-based hardware, or adapted to meet development needs.

The emergence of AI-defined vehicles adds a new layer of urgency to this compute transformation. These vehicles depend on low-latency, edge-based AI processing across power-efficient, high-performance platforms.

From ADAS to conversational IVI, each AI workload demands real-time coordination across heterogeneous compute units like CPUs, GPUs, and accelerators. Arm’s AE technologies enable this compute consistency with integrated safety and security features aligned to ISO 26262 and ISO/SAE 21434 standards.

Maintaining Performance and Reliability

Conducting comprehensive reliability tests, such as vibration, thermal cycling, and humidity testing, helps ensure that components can withstand various operational stresses. Arm Software Test Libraries (STL) offer a suite of diagnostic tests to verify the correct operation and integrity of Arm-based microcontrollers (MCUs), helping to ensure that faults are detected early. These libraries are especially useful to help enhance the reliability of safety-related applications, such as ADAS.

Meanwhile, Arm Compiler for Embedded FuSA is an ISO 26262-qualified toolchain, which helps to boost the performance, code density, architectural accuracy, and safety of software applications within the vehicle.

You can learn about Arm's reliability testing tools here.

Regularly updating vehicle software and continuously monitoring system performance can also help maintain reliability. Arm solutions support OTA updates, allowing vehicle manufacturers and Tier-1 suppliers to fix bugs, improve functionality, and enhance security without requiring physical access to the vehicle.

For instance, the Arm Reference Design-1 AE, or RD-1 AE, introduces the concept of a high-performance Arm Neoverse V3AE application processor for primary compute, augmented with an Arm Cortex-R82AE based Safety Island for scenarios where additional system safety monitoring is required. The system additionally includes a runtime security engine (RSE) used for the secure boot of the system elements and the runtime secure services.

Arm also offers different classes of processor, each with a broad range of capabilities, specifically designed to address the needs of a variety of automotive applications.

Scalable Silicon: Top strategies to ensure security across vehicle configurations.

Today’s vehicles increasingly rely on centralized compute architectures to handle diverse functions such as powertrain, IVI, ADAS. Therefore, the trend toward modular and scalable designs is becoming more prevalent, allowing automotive OEMs to develop platforms that can be easily adapted to different vehicle segments and configurations.

Developing execution technologies are critical for preventing control-flow attacks and ensuring that only authenticated code can execute. Arm processors incorporate security features like Pointer Authentication and Branch Target Identification (PACBTI), which help prevent such attacks by verifying the integrity of the code before it runs. These technologies are essential for maintaining the integrity of the system, especially in a highly connected automotive environment.

The following video helps explain how Arm is revolutionizing the automotive industry with robust, reliable system architecture and integrated security solutions.

End-to-End Security with Arm

Securing the world’s data is one of the greatest technology challenges during the next decade of compute. It is a challenge that can only be tackled and scaled with collaboration across the entire ecosystem. Arm is at the forefront of continued security research and industry collaborations to democratize the development of security technologies for AI-enabled SDVs.

One of the most challenging aspects of securing devices and systems revolves around software. Arm’s MTE and RME security technologies isolate different software components, which in turn helps prevent malicious code from affecting critical system functions.

MTE enables the detection of memory safety violations, such as buffer overflows, by tagging memory allocations and checking these tags during memory accesses. This helps to prevent common vulnerabilities that could be exploited by attackers.
RME provides a secure execution environment that isolates sensitive data and code from the rest of the system, helping to ensure that even if the main operating system is compromised, critical functions remain protected.

Standard security APIs are crucial for ensuring interoperability and security across different platforms.

PSA Certified for Security

PSA Certified provides a framework for implementing security best practices and standardized APIs, helping to ensure that all components in the automotive ecosystem adhere to the same high-security standards. As the number of automotive and industrial applications with functional safety requirements has grown significantly, internationally agreed standards are now in place to help ensure systems achieve the required levels of reliability and safety integrity.

Arm Software Test Libraries

Arm Software Test Libraries (STL) are software-based safety mechanisms which provide diagnostic capabilities for permanent faults. Based on the level of diagnostic coverage that can be achieved, they are ideal for lower safety integrity requirements, such as ISO 26262 ASIL B automotive applications and IEC 61508 SIL 2 industrial systems. In addition, they are typically less complex to deploy than built-in self tests (BIST) and use less silicon area and power.

Learn more about Arm Software Test Libraries here.

Integrate Arm STLs for a Safe Automotive Experience

Hardware must work correctly across high-reliability automotive systems. Common methods to check the integrity of the silicon include BIST for memory (MBIST) and CPU logic (LBIST). For the highest safety standards, an evolved industry standard is Dual-Core Lock-Step (DCLS).

In addition to hardware integrity checks, software tools like the Arm STL offer an efficient and cost-effective way to enhance system reliability. STL can be executed within a running application environment to help ensure smooth operation in automotive systems and boost the reliability of safety-related applications, all while meeting stringent ISO26262 standards.

AI-defined vehicles will further amplify the need for scalable and secure silicon across multiple vehicle platforms.

As AI workloads become more diverse and personalized, from driver monitoring to predictive safety, OEMs must ensure that security is enforced across every layer of the compute stack. This requires compute platforms that deliver both flexibility and robust isolation.

Arm’s AE technologies allow OEMs to tailor AI performance and security needs to individual vehicle platforms, without compromising safety or time to market.

AUTOSAR Integration

The integration of Arm STL into Classic AUTOSAR, which is a standardized framework for real-time deterministic automotive software, brings various benefits, including:

Robust fault detection: Arm STL offers a comprehensive suite of diagnostic tests to verify the correct operation and integrity of Arm-based MCUs, including those built on Arm Cortex-M and Arm Cortex-R. This helps ensure that faults are detected early, preventing potential failures in critical automotive functions.
Improved reliability: By integrating STL, developers can enhance the reliability of ECUs, helping to ensure that the hardware operates correctly before the application runs. This is crucial for meeting stringent ASIL requirements.
Flexible integration: This flexibility allows developers to choose the best approach for their specific application needs.

Learn more on how to integrate Arm STL into Classic AUTOSAR.

Meanwhile, microcontroller hypervisors enable the consolidation of multiple applications into a single ECU, which helps ensure freedom from interference and supports mixed criticality systems. Integrating Arm STL into this architecture offers several advantages:

Enhanced fault mitigation: The hypervisor's ability to separate applications is crucial for maintaining system integrity. By running STL at the hypervisor level (EL2), developers can achieve high diagnostic coverage, helping to ensure that faults are detected and mitigated effectively.
Scalable and flexible systems: The integration strategies described in the white paper below allow for flexible and scalable system designs. Whether running STL during boot-time or periodically during operation, developers can tailor the integration to meet their specific safety and performance requirements.
Support for mixed-criticality systems: The ability to run multiple software stacks on a single ECU, with assured spatial and temporal separation, is a game-changer. This supports the development of systems where applications of varying safety levels can coexist without compromising overall system integrity.

Learn more on integrating Arm STL into microcontroller hypervisors.

The Transformative Role of Chiplets in the Automotive Industry

Adopting a modular and scalable architecture helps enhance the vehicle’s software and hardware adaptability. Arm Zena CSS enables OEMs to scale AI capability across car tiers and domains — from IVI and ADAS to central compute. It reduces silicon engineering effort by up to 20% and software porting time by up to 30%, enabling faster development, lower costs, and consistent performance from edge to cloud.

This modular approach allows OEMs to customize and scale their ADAS solutions to meet specific regional requirements and evolving global regulations.

Additionally, chiplets offer numerous intriguing SoC design possibilities for the ecosystem. Through our partner ecosystem, Arm is delivering a versatile compute platform to support the expanding chiplets ecosystem.

Our solution stands out by offering multiple reusable IP components that can be integrated into larger systems by partners, as showcased across our new Automotive Enhanced IP technologies.

In this video, we explain what are chiplets - a multi-chip modules that allow process nodes to be optimized for subsystems rather than expensive process nodes to create a flexible path to custom silicon that helps reduce costs and allows existing components to become building blocks for your custom chips thanks to a scalable, modular design.

Vehicle Security: Protecting architectures from external threats and vulnerabilities.

To protect vehicle architectures from physical security threats, vehicle manufacturers can implement robust hardware-based security measures. This includes integrating Hardware Security Modules (HSMs) and Trusted Execution Environments (TEEs) into vehicle systems. HSMs provide a secure environment for cryptographic operations, helping to ensure that sensitive data is protected from tampering and unauthorized access. TEEs, on the other hand, create isolated environments within the main processor, safeguarding critical operations from potential threats.

Physical Security Measures

Arm Automotive Enhanced (AE) IP processors, such as those based on the Armv9 architecture, incorporate these advanced security features to provide a strong foundation for protecting vehicle systems. Arm’s approach to automotive security includes defensive execution technologies, such as Pointer Authentication and Branch Target Identification (PACBTI), which help protect against control-flow attacks. These technologies help ensure that the execution of code is secure and resistant to tampering.

Meanwhile, Arm, alongside other industry leaders such as AWS, Continental, BOSCH and Red Hat, founded SOAFEE, which is driving automotive industry standardization to reduce software complexity. Since then, a vast ecosystem has grown, using SOAFEE methodologies for a multivehicle platform, single software approach.

SOAFEE members have already created a brand-new ecosystem of software solutions that are critical to enabling Arm Compute Subsystems (CSS) for Automotive in 2025 through delivering software consistency to support the silicon development and deployment process.

Now, SOAFEE.next is bringing together reference designs for hardware and software by aligning SOAFEE-based software solutions with the Arm Reference Design-1 AE hardware solution for automotive which leverages the AI, security, safety, and virtualization capabilities of the latest Armv9 technology.

This alignment is key, and one of the biggest achievements of the SOAFEE initiative to date, as it enables mixed critical development for hardware and software for a variety of functional safety and application workloads. This means different software-defined functions in SDVs can run at different levels of safety criticality using the SOAFEE architecture.

Learn more on how Arm is pioneering the SDV revolution.

Arm MTE and RME technologies isolate different software components, which in turn helps to prevent malicious code from affecting critical system functions. MTE is an extension to the Arm architecture which helps detect errors in software use of memory, such as out-of-bounds accesses.

Learn more about internal hardware security modules here.

Meanwhile, Arm’s suite of security services, including secure boot and run time security, help ensure that the system starts and operates securely. Developers can also use standard security APIs to help ensure interoperability and ease of integration, creating a consistent security framework across different automotive platforms.

Learn more about external hardware security module here.

Supply Chain Security Measures

Much of the software stack in vehicles is proprietary, so vulnerabilities in the code are less well-known and harder to find. Similar software that is used across adjacent markets, like consumer technology and IoT, could be leveraged and exploited in the vehicle, leading to further vulnerabilities.

Arm's support for maintaining a Software Bill of Materials (SBOM) helps automakers track and manage software components throughout the supply chain so vulnerabilities can be quickly identified and addressed. This proactive approach helps mitigate risks and enhances the overall security of vehicle architectures.

One of the primary challenges in secure firmware updates is ensuring that only authenticated and authorized updates are applied to the system. Unauthorized updates can introduce vulnerabilities or malicious code and compromise the security and functionality of the device. This requires robust mechanisms to verify the authenticity of the update source.

Secure boot processes and firmware updates are essential for protecting vehicle systems from unauthorized modifications. Secure boot helps ensure that only trusted software is executed during the startup process to prevent malicious code from being loaded.

Arm TrustZone technology has a critical component in enhancing the security of vehicle systems. It provides a hardware-based root of trust, which is essential for implementing secure boot processes and secure firmware updates. At the core of TrustZone technology is the concept of creating two distinct execution environments: the secure world and the non-secure world. This hardware-enforced isolation helps ensure that sensitive operations and data are protected from potentially malicious software running in the non-secure world.

Arm has also introduced Dynamic TrustZone, an innovative extension that enhances the flexibility and efficiency of TrustZone technology. Dynamic TrustZone uses RME to allocate memory between the secure and non-secure worlds at runtime. This capability is particularly useful for managing complex workloads and maximizing efficient use of resources in modern vehicle systems.

Automotive Software: Best strategies to complex diverse software offerings with real-time processing.

The automotive industry is moving towards modular and scalable architectures to manage the increasing complexity of software workloads. This trend is driven by the need to consolidate multiple functions into fewer, more powerful units, reducing the number of ECUs and simplifying the overall system.

Arm zonal architectures consolidate processing power into fewer, more powerful units, allowing for easier updates and maintenance. This modularity helps enhance the ability to support diverse software applications by enabling individual modules to be updated or replaced without affecting the entire system.

This shift helps OEMs protect software investments and enables post-production monetization across different car segments. Additionally, the evolution of the automotive EE-architecture to zonal architecture with central compute supports the trend toward further software integration in cars. This helps minimize the overall number of ECUs in the car, which reflects the requirements of the AI-enabled SDV for ongoing software updates and maintenance.

Learn more about zonal and centralized compute architecture.

The Arm Automotive Enhanced (AE) IP suite, including the Arm Cortex-A and Arm Cortex-R series of processors, provides the necessary performance and flexibility to support modular and scalable architectures. These processors are designed to handle the complex workloads of modern vehicles, helping to ensure efficient and reliable operation.

For automotive embedded-software development, the right tools must comply with safety and security standards to evaluate, prototype, and test software.

As software continues to define the driving experience, AI-defined vehicles are emerging as the next evolution, pushing compute performance and safety demands even further. These vehicles will run diverse AI workloads at the edge, from predictive ADAS to personalized IVI systems, requiring real-time responsiveness and low-power efficiency.

To meet these needs, vehicle platforms must support heterogeneous computing across CPUs, GPUs, and accelerators while maintaining functional safety and cybersecurity standards like ISO 26262 and ISO/SAE 21434. Arm’s Automotive Enhanced (AE) technologies deliver the scalable compute foundation required to run complex AI applications efficiently and securely, across vehicle segments, workloads, and evolving software stacks.

Get started with automotive software tools from Arm.

RTOS for Time-Sensitive Tasks

The use of real-time operating systems (RTOS) is becoming increasingly important in automotive applications to help ensure that time-sensitive tasks are executed within strict deadlines. This is crucial for safety-critical applications such as collision avoidance and braking systems.

RTOS provides deterministic behavior, which is essential for maintaining the reliability and safety of automotive systems. The Arm Cortex-R82AE processor, for example, delivers high performance and real-time processing capabilities, making it ideal for safety-critical applications. This processor supports functional safety standards and provides the deterministic behavior required for real-time processing.

The Cortex-R82AE is a 64-bit Cortex-R processor that blends the benefits of deterministic execution with advanced processor technologies for rich software stacks, including coherency and virtualization. It is capable of providing high single-thread performance in multicore, multicluster systems as the primary processor, or as a real-time safety controller with simplified integration with applications processors in heterogeneous systems.

Arm Cortex-R82AE enables the flexible deployment of different software stacks. It can support both real-time software stacks, such as those deployed using Classic AUTOSAR in many established control applications, as well as the ability to run rich software stacks, such as Linux, for service-based functions. This software flexibility helps users readily change the software partitioning used in an automotive computing system as zonal controller requirements evolve.

Learn more about the advantages of Arm Cortex-R82AE here.

AI-defined Compute: Synchronizing hardware and software in centralized systems.

The automotive industry is rapidly transitioning toward AI-enabled SDVs which rely on centralized compute systems to control most of their functionality, from engine performance to infotainment systems.

Transitioning to a centralized architecture involves consolidating multiple ECUs into fewer, more powerful central units. This shift helps reduce the complexity of vehicle systems, lower costs, and decrease the overall weight of the vehicle, which can improve fuel efficiency and performance. Centralized systems also help enhance scalability, allowing OEMs to integrate new features and technologies more easily than ever.

Arm high-performance processors provide the computational power and flexibility needed for these centralized systems, helping to ensure they can handle the increased processing demands of modern SDVs.

Arm and NXP offer insights into new ways to approach real-time software computing strategies using the latest Arm automotive technologies in this on-demand webinar.

Transform Vehicle Systems With Decoupling and Virtualization

Decoupling software from hardware is essential for maintaining flexibility and future-proofing vehicle systems. This approach allows OEMs to update software independently of hardware changes, facilitating OTA updates and helping to reduceg the need for costly physical recalls.

By standardizing interfaces between software stacks and underlying hardware, Arm's support for the VirtIO framework enables seamless integration and updates. This decoupling not only simplifies the update process, but also allows for the quicker deployment of new features and improvements.

Panasonic Automotive Systems and Arm have formed a strategic partnership to standardize the architecture for SDVs. This collaboration aims to decouple automotive software development from hardware using the VirtIO framework, helping to improve flexibility and accelerate development cycles. The partnership also focuses on creating a unified software stack and standardizing zonal architectures to help improve efficiency and performance in modern vehicles.

Robust Virtualization for Performance and Security

Virtualization is key for managing multiple functions on a single hardware platform. It helps ensure that different software components can run securely and efficiently, without interfering with each other. Arm solutions, such as the Arm Mali-G78AE GPU, support robust virtualization and allow OEMs to partition resources effectively.

This is crucial for maintaining performance and security in a centralized compute system as it enables the isolation of critical functions and the efficient allocation of processing power.

Explore the full details here.

Designed specifically for targeting both safety and flexibility, a single multicore GPU can be divided into separate hardware partitions for different workloads. A single SoC can be designed and then configured at boot time into various domains or deployment use cases. Each domain can operate as a separate GPU. With flexible partitioning, these domains can be shown to be separate and isolated from each other.

For cybersecurity, integrating Arm Automotive Enhanced (AE) technologies with built-in security mechanisms, like Arm TrustZone, provides robust protection against cyber threats. These include hardware-enforced isolation for critical code and data, helping to ensure that sensitive operations are protected from malicious attacks.

Additionally, Arm follows state-of-the-art product security practices and standards, like ISO/SAE 21434, to help ensure that security risks are managed during the idealization, development, and post-development of all products. Arm provides a set of supporting security materials for automotive partners to simplify the integration of our off-the-shelf components into ISO/SAE 21434-compliant designs.

Looking ahead, these centralized compute systems will lay the groundwork for the AI-defined vehicles of the future. As vehicles transition into intelligent platforms powered by edge AI, the ability to handle real-time, high-volume workloads across secure and isolated environments becomes even more essential.

AI-defined vehicles will be built on scalable, high-performance platforms like Arm's AE technologies to deliver the compute and virtualization needed to support AI-driven decision-making, advanced personalization, and autonomous capabilities, all within a tightly integrated hardware-software framework.

While AI-defined vehicles bring new demands, they will be built on the centralized compute and virtualization foundations established by SDVs, making this cloud to car infrastructure essential for long-term scalability and adaptability.

AI-Driven SDVs: Ensuring security and reliable OTA updates in SDVs.

With a billion lines of code expected in the AI-enabled software-defined vehicles (SDVs) of the future and substantial increases in connectivity, the automotive attack surface continues to grow and evolve. One challenge with SDVs is keeping interiors updated and matching the pace of iteration and innovation of the vehicle’s software. Nobody wants to drive a vehicle that feels futuristic in its mechanics, but looks vintage in its interior.

Additionally, with the shift toward alternative powertrains and autonomous driving, designers can reimagine traditional vehicle layouts, introducing sleeker aerodynamics, innovative lighting systems, and customizable exteriors. This can be further enhanced as the possibilities from 3D printing continue to evolve and become more mainstream, with OTA software updates supporting updated and new exteriors from 3D printing techniques.

Our latest Dummies Guide helps you understand the SDV world with important insights into technologies and systems and their impact on the future of transportation. Click here to learn more.

Arm AE Technologies for Automotive

Arm Automotive Enhanced (AE) technologies are designed to meet increased levels of performance, safety, and security for SDVs. The new range of Armv9-based AE processors include the very latest security and software features.

Arm continues to be at the forefront of providing security-focused architecture features, empowering its technology ecosystem to protect businesses, individuals, and devices. This expertise is extensively deployed in the automotive industry, where Arm collaborates with the ecosystem to deliver and activate the latest Armv9 architecture security features, alongside continuous efforts on standards and open-source software.

The latest Arm AE processors incorporate key Arm defensive execution technologies and architecture features designed to mitigate the risks posed by vulnerable or malicious software. Technologies such as Pointer Authentication (PAC), Branch Target Identification (BTI), and Memory Tagging Extension (MTE) help safeguard the integrity of software control flow and reduce the impact of memory safety bugs.

This is particularly crucial for the automotive sector, which still relies on a substantial amount of legacy code written in memory-unsafe languages like C, that can be ported into SDVs.

Support for Latest Security Practices and Standards

Arm adheres to state-of-the-art product security practices and standards, such as ISO/SAE 21434, to manage security risks throughout the ideation, development, and post-development phases of all products. Arm provides supporting security materials to automotive partners to help simplify the integration of off-the-shelf components into ISO/SAE 21434-compliant designs.

Scalable isolation technologies are also integral to the Arm architecture, enabling the segregation of diverse workloads with minimal performance impact. This is essential in the automotive industry, where there is a need to run mutually distrusted software components from various sources on the same computing platform. Technologies like Arm TrustZone, S-EL2, and RME help enforce well-defined trust boundaries to support this requirement.

Security, however, extends beyond hardware. Arm supports the software ecosystem through frameworks and API solutions, helping to ensure the effective deployment of architecture features. Arm co-founded and continues to contribute to standard security APIs, such as the PSA Certified Crypto API, which serves as a contract between firmware developers and hardware vendors. This allows developers to focus on firmware design without needing to understand proprietary hardware rules for each new integration. For hardware vendors, standard APIs remove barriers to entry, allowing them to concentrate on valuable commercial differentiation.

In this exclusive interview, Magnus Östberg, Chief Software Officer at Mercedes-Benz, sits down with Dipti Vachani, Senior Vice President and General Manager for Automotive at Arm, to explore the key role of software in transforming the automotive industry.

As the industry progresses toward AI-defined vehicles, OTA updates will no longer simply patch features, but they will continuously evolve a vehicle’s intelligence. From AI-enhanced collision prediction to adaptive IVI systems, AI-defined vehicles will require dynamic software updates that enhance both safety and experience without introducing security vulnerabilities.

Arm’s alignment with ISO/SAE 21434 and support for standard APIs ensures that these ongoing updates can be delivered securely and efficiently, even in highly connected, AI-driven environments.

AI-defined vehicles will amplify the need for secure, reliable OTA updates, but these capabilities will be built on the cloud-to-car infrastructure pioneered by SDVs.

Environmental Parity

Arm’s ISA parity, also known as "environmental parity," refers to the alignment of the Instruction Set Architecture (ISA) used in the cloud with the ISA used at the edge, such as applications and devices in the actual vehicles. This parity is achieved thanks to the success of Arm Neoverse in the cloud and the adoption of new Arm AE solutions in vehicles, both built on the Armv9 architecture.

Leading cloud platforms like AWS, Ampere, Google Cloud, Microsoft Azure, and Oracle now offer Arm-based compute instances to the automotive industry. For example, for software developers, this means that development work done in the cloud on Arm Neoverse-based AWS Graviton have almost exact parity with automotive applications built on the new Arm Cortex-A720AE.

This allows software to be built and tested in the cloud and then deployed at the edge in vehicles with the same binaries, requiring only a recompile. Identical toolchains help eliminate the need for cross-compilation or different sets of drivers and speed up the development process.

You can learn more about ISA Parity here.

SOAFEE for Embedded Edge provides a robust framework to help address the challenges of seamlessly integrating virtual systems in vehicles. The SOAFEE architecture supports the development of SDVs through decoupling hardware and software, which allows for a more flexible and scalable integration of different systems.

The SOAFEE cloud-native approach helps in the development of vehicle applications in a virtual environment before they are validated in real-world conditions. This not only helps reduce development time, but also helps ensure that applications can be seamlessly integrated into various vehicle architectures.

Additionally, SOAFEE support for mixed-criticality workloads helps ensure that both safety-critical and non-safety-critical applications can coexist and operate reliability within the same system.

Learn more about SOAFEE through GitLab.

AI-defined Vehicle Development: Validating real-time performance and cross-silicon portability.

The SDV revolution is already underway and set to transform the automotive industry. This transformation requires industry-wide collaboration, standardization, and the embrace of new software methodologies and architectural approaches.

SDVs require advanced computing power, effectively transforming them into datacenters on wheels. This metamorphosis calls for the meticulous integration and management of both hardware and software components. To achieve optimal performance, it is essential to select suitable hardware components, while monitoring and maintaining software components for safety and reliability.

SDVs demand the seamless integration of hardware and software components, with a focus on safety and performance. Prioritizing safety and reliability features is crucial to protecting occupants and other road users. The following aspects of design are essential to meet the growing demand of the SDV and the massive computation requirements and data it must process:

Efficient but well partitioned hardware and software integration.
Hardware that enables seamless software portability and updates.
Continuously monitored software components for safety and reliability.
Regularly updated and upgraded software for enhanced performance.

Arm is at the forefront of the SDV revolution, offering essential technologies for automotive applications to make SDVs a reality. Our AE IP portfolio, which includes Arm Cortex-A, Arm Cortex-M, Arm Cortex R CPUs, Mali GPUs, and ISPs, forms the backbone of many automotive computing solutions for SDVs.

Key innovations and initiatives by Arm for SDVs include:

High-performance processors: The latest Armv9-based AE IP processors enhance security and performance in SDVs, featuring advanced security features like Branch Target Identification (BTI), Pointer Authentication (PAC), and Memory Tagging Extension (MTE).
Scalable platforms: Arm scalable platforms support a wide range of automotive applications, from advanced driver assistance systems (ADAS) to in-vehicle infotainment (IVI), enabling seamless integration and upgradability.
Ecosystem collaborations and initiatives: Arm collaborates with numerous automotive manufacturers and technology companies to advance SDV technologies. These partnerships are crucial for accelerating the development and deployment of SDVs.
Comprehensive solutions: Arm is working with a wide range of partners, including Amazon Web Services (AWS) and Tata Technologies, to deliver comprehensive solutions for SDVs. These collaborations focus on integrating advanced technologies and ensuring that SDVs meet the highest standards of performance, safety, and security.
Virtual platforms: Arm and our partners including Cadence, Corellium, and Siemens have developed virtual platforms that allow for the early evaluation and development of software without the need for physical silicon. This helps significantly reduce the time-to-market for new automotive technologies.

Meanwhile, Arm, alongside other industry leaders such as AWS, Continental, BOSCH and Red Hat, founded SOAFEE – Scalable Open Architecture for Embedded Edge in 2021, which is driving automotive industry standardisation to reduce software complexity. Since then, a vast ecosystem has grown, using SOAFEE methodologies, which enable vehicle manufacturers to implement a multi-vehicle platform, single software approach.

This SOAFEE membership has already created a brand-new ecosystem of software solutions that will be critical to enabling  Arm Zena Compute Subsystem (CSS)  through delivering software consistency to support the silicon development and deployment process.

Moreover, SOAFEE.next is bringing together reference designs for hardware and software by aligning SOAFEE-based software solutions with the Arm Reference Design-1 AE hardware solution for automotive which leverages the AI, security, safety, and virtualization capabilities of the latest Armv9 technology.

The need for cross-silicon portability becomes even more pressing in the context of AI-defined vehicles. These will rely on AI models trained in the cloud and deployed at the edge for lower latency. With ISA parity across Arm Neoverse-based cloud environments and Arm AE-powered compute in the vehicle, developers can build, validate, and deploy AI workloads once, reducing friction while meeting real-time demands. This cloud-to-car consistency is critical to enabling the adaptive, learning-based functionality that defines the AI-defined vehicle era.

Learn more on how Arm is pioneering the SDV revolution

SDV Safety: Complying with functional safety standards while managing OTA updates.

Changing consumer demands for safer, smarter, and more connected vehicles has resulted in an unprecedented evolution of the automotive industry. This is coupled with the rise of autonomous driving, advanced driver assistance systems (ADAS), and electric vehicles. As a result, the overall complexity of automotive systems has increased significantly, bringing new safety challenges that must be addressed to protect consumers, while ensuring an optimal driving experience.

Arm’s approach to safety and performance delivers:

Enhanced performance and efficiency: High-performance processors from the Arm Automotive Enhanced (AE) portfolio.
Seamless integration: A comprehensive support package, including Arm software test libraries (STL), tools, and compilers.
Accelerated development: Virtual platforms for pre-silicon software development and testing, reducing time-to-market.
Compliance with safety standards: Adherence to UNECE GSR, ISO 26262, ISO 21448 (SOTIF), and ISO 21434 standards across Arm Safety Ready solutions and AE processors.
Robust ecosystem: The SOAFEE initiative helps reduce software complexity and enables a multivehicle platform, single software approach.

Functional safety helps ensure that vehicles respond correctly to their inputs, thereby preventing malfunctions and safeguarding passengers. A vital part of this is building products for the automotive industry that are functionally safe by complying with the relevant standards for that specific application.

For example, an ADAS solution must adhere to UNECE GSR and standards like ISO 26262, ISO 21448 (SOTIF), and ISO 21434, ensuring correlation between safety and security. Arm’s adherence to industry standards like ISO 26262 provides a strategic framework for the design, testing, and verification of ADAS systems.

AI-defined vehicles will bring an additional layer of complexity to safety validation. They will incorporate deep learning models that influence real-time driving behavior, emotion recognition, and predictive ADAS features, requiring not only functional safety compliance but also runtime adaptability.

AI-defined vehicles will rely on the safety frameworks, compliance standards, and OTA validation systems that are being matured within today’s SDVs, making them deeply interconnected stages in vehicle evolution.

To support this, Arm’s safety-ready compute platforms, combined with continuous software validation in virtual environments, enable developers to test and optimize AI features and workloads under safety-critical constraints long before deployment.

Discover key considerations for automotive functional safety.

Virtual Platforms

To address the complex, expensive, and time-consuming task of developing, verifying, and validating various SDV functions, Arm has collaborated with partners, such as AWS, Siemens, and Corellium, to offer a new breed of virtual platforms.

The availability of such development environments for Arm’s latest AE solutions means that software can be developed and tested before the physical silicon is ready, helping to reduce development time and accelerate time-to-market.

Meanwhile, Arm, alongside other industry leaders such as AWS, Continental, BOSCH and Red Hat, founded SOAFEE, which is driving automotive industry standardization to help reduce software complexity. Today, a vast ecosystem uses SOAFEE methodologies to implement a multivehicle platform, single software approach.

Arm, in collaboration with Codethink, has tackled a crucial part of this integration puzzle by incorporating Arm STL into live Arm Cortex-A application processor software stacks. This specifically addresses the integration at Exceptional Level 3 (EL3) within Trusted Firmware-A (TF-A) for runtime utilization, which allows developers to enhance the integrity and reliability of their systems. Read more here.

Arm is working to address the challenges of software portability, cloud-native mixed-criticality, and seamless development. Learn more.