Software-Defined Vehicle Magazine – June / July 2025

The automotive industry is shifting from a hardware-centric model to a software-centric one. This emerging paradigm calls for increasingly rapid development cycles from automakers. From new in-cabin experiences and advanced driver assistance systems (ADAS) to cyber security and data management, the journey towards the software-defined vehicle (SDV) is rewriting the rulebook on automotive innovation.

IBM is building on its roots in computing to help automakers modernise their operations in response to the new industry clock speed. Drawing on more than a century of technology expertise, it now finds itself on the vanguard of new mobility development. “In some cases we are providing the IT infrastructure for automotive projects—a very traditional IBM play. In others, we directly supply something that goes into the vehicle, like our Red Hat In-Vehicle Operating System,” says Hans Windpassinger, Principal Client Engagement within IBM’s Global Manufacturing Industries team. “Whether it’s technology, consulting or support services, we are helping automotive customers on their journey towards SDV.”

IT infrastructure lags AI and compute

ADAS is growing increasingly clever thanks to progress within artificial intelligence (AI), but training AI models requires vast amounts of data. Many developers find themselves confronted with petabytes of data spread across various regions, and managing that data is a huge challenge. “It’s not unusual for OEMs to collect up to 40 or even 50 petabytes of data,” he says. “They cannot simply make a copy or put it somewhere. You have to manage it effectively, and managing huge amounts of data has always been a strength of IBM.”

Continental served as a case study on the effectiveness of IBM’s IT infrastructure. The supplier was regularly receiving more than 100 terabytes of data from a test fleet of automated vehicles. With the help of Nvidia GPU computing, this data was ingested, processed, selected, annotated, and used for training and validation of the system. “Continental invested in these very data-hungry GPUs from Nvidia and then realised that its existing infrastructure cannot really feed them,” notes Windpassinger.

What the supplier needed was something faster that could provide the necessary bandwidth and random access. For this it turned to BMW’s Elastic Storage Server (ESS) and found that it could enable more AI experiments within the same time compared to the legacy system. “It could get 14 times more out of the Nvidia DGX systems with our ESS,” emphasises Windpassinger. “This is an incredible factor, which shows how important IT infrastructure is for enabling efficient work from data scientists and leveraging the GPU power available today.”

The right IT infrastructure helps developers get the most out of the computing power now available The open source SDV

With the move towards SDV, the in-vehicle operating system (OS) becomes increasingly important. It acts as the vehicle’s brain, serving as the core software platform and coordinating multiple electronic control units and functions. The global automotive OS market is poised to grow in value from US$11.83bn in 2023 to US$30bn by 2030, according to Grand View Research. There are a number of solutions on the market today, but IBM is hopeful that its new Linux-based offering could propel the industry towards the open source SDV.

The Red Hat In-Vehicle Operating System comes from IBM’s Red Hat subsidiary and applies Linux to safety-critical automotive systems. In theory, it accelerates development time, reduces costs, and opens the door to new services and revenue streams. The idea is that the huge existing talent pool of developers competent in Linux will make it easier for automakers to find the necessary talent and innovate faster. “More talent means faster innovation,” suggests Red Hat.

Automotive certification remains a work in progress. So far, the Linux math library, a key part of the Red Hat In-Vehicle OS, has achieved ISO 26262 ASIL-B functional safety certification. It has also demonstrated that the required mixed criticality configuration can reliably meet ASIL-B requirements. “By pairing edge computing with the proven reliability of Linux, we’re building a foundation that not only meets critical safety standards but also enables automakers to deliver personalised, connected, and adaptive driving experiences at scale,” comments Francis Chow, Vice President and General Manager, In-Vehicle Operating System and Edge, Red Hat.

While Windpassinger isn’t able to share specific timelines, he adds: “We really look forward to the time when we have the Red Hat In-Vehicle OS completely certified.”

Security

Software is extending into almost every aspect of the vehicle and the development process, meaning cyber security needs to do the same. Guidehouse Insights expects the global automotive cyber security solutions market to see revenue of more than US$445bn by 2031. “Security is a really important element within SDV. As the attack surface grows, we also see the number of vulnerabilities—and incidents—growing,” Windpassinger observes. Securing both hardware and software while conforming to regulatory requirements and maintaining customer expectations has put traditional ways of working under pressure. In many cases, automakers are turning to outside expertise.

IBM has a dedicated security unit, X-Force Red, with the stated mission of “hacking anything to secure everything”. One year after its 2016 launch X-Force Red added a division specifically focused on automotive. Here, security experts test vehicles and components for vulnerabilities and work with manufacturers to fix them. “These guys are the good hackers. They hack systems based on an order from a customer, and they always find something. Nothing’s perfect.”

Looking ahead, Windpassinger sees considerable scope to improve automotive cyber security with greater use of AI. Within the vehicle security operation centre, AI can be used to find patterns and to better identify potential incidents. He also flags threat analysis risk assessments and penetration testing as areas that could be improved with AI.

X-Force Red is a dedicated cyber security business with a special unit just for automotive Growing in importance

The rise of software’s role within automotive is blurring the lines of industry players and relationships. While IBM didn’t start off as an automotive company, its expertise is becoming increasingly relevant to the industry. As well as the areas touched on above, the company also offers automotive players its expertise on large language models (LLMs), algorithms that are particularly good with human language. “Our own LLMs, called Granite, are small, effective and powerful. They are perfect candidates for embedding into vehicles,” he says. The latest addition to the line-up, Granite Family Version 3.0, is currently provided on the Qualcomm AI Hub, meaning it can be executed and deployed on the ubiquitous Snapdragon platform.

The company’s connected vehicle platform, IBM IoT Connected Vehicle Insights, helps automakers analyse data from connected vehicles and turn it into actionable insights. It has already been adopted by big names such as Suzuki in Japan and its Indian unit Maruti Suzuki. The SDV portfolio also includes the IBM Engineering Lifecycle Management platform for engineers, helping with modelling and compliance.

“With the rise of SDV, automotive is definitely growing in importance for us and our customers,” concludes Windpassinger. “Despite the current challenges, OEMs and Tier 1s are pushing ahead with their R&D.” And with advances like these, IBM is becoming increasingly efficient and responsive to the needs of software-defined mobility.