Intel Pulls the Plug on Automotive

In June 2025, it was first reported and later confirmed by Intel that it will be winding down its automotive business within the client computing group. This move will bring to a close decades of supporting the automotive industry, and, in particular, an intense 10-year period that saw Intel investing in the industry across various domains. Highlights include the US$15 billion acquisition of Advanced Driver-Assistance Systems (ADAS) and autonomous vehicle software developer, Mobileye, which was later floated through an Initial Public Offering (IPO) in 2022, and the more recent acquisition of Electric Vehicle (EV) electronics specialist, Silicon Mobility, in 2024. With automotive megatrends calling for ubiquitous Artificial Intelligence (AI) and sever-ized Electrical/Electronic (E/E) networks, and Intel making investments and acquisitions across multiple key automotive domains, Intel pioneered an AI Everywhere approach that emphasized its ability to bring the best of its experience in the Artificial Intelligence (AI) Personal Computer (PC) and data center to the software-defined vehicle, at the same time empowering automaker customization through chiplets.

However, Intel once again finds itself withdrawing from a market in which it should have been a major player. While intel's limited revenue in the automotive space ultimately means that few individual automakers will be affected, and it should be noted that Intel has committed to supporting customers with a smooth transition, the sudden departure of a semiconductor giant from the automotive space only months after impressive showcases at CES and Auto Shanghai 2025 is bound to set some alarm bells ringing within many automakers.

For most of their history, automotive Original Equipment Manufacturers (OEMs) were able to benefit from a captive supply chain. In both their innovation and their manufacturing, automakers were supported by a network of automotive-native or automotive-oriented component supplies and system integrators, providing OEMs with the best of both worlds. As automotive native suppliers, they had a proven commitment to the market and their ability to support OEMs in the long term and the long automotive life cycle, while at the same time, the presence of multiple competitive suppliers kept prices in check, and gave automakers the redundancy of an alternative source if needed.

However, as automakers felt the pressure of keeping up with consumer expectations for digital experiences in the cabin, and ever-aware of the potential competitive threat in the long term of driverless mobility services originating from outside the conventional automotive supply chain, it became clear to automakers that they must look beyond their conventional and captive supply chain to secure the key enabling technologies behind rich infotainment, autonomous driving, and software-defined vehicles. If they were going to mimic smartphone experiences in the cabin, it would be by partnering with smartphone silicon suppliers, and it they were to successfully deploy AI workloads on the road, it would be by partnering with the same silicon suppliers that accelerate AI in consumer electronics and the data center.

Nevertheless, many automakers made these vital strategic partnerships with trepidation, fearing that the same semiconductor giants that drove technology innovation in smartphones, laptops, and data centers would not understand the unique requirements of the automotive market. Some decision makers within OEMs were skeptical concerning the long-term commitment of semiconductor newcomers to the automotive market, They feared that, as the smartphone and PC markets began to saturate, silicon giants were keen to demonstrate to investors their ability to continue to grow through new markets, such as automotive, but that as the realities of a complex market that ships, at best, 90 million units a year began to set in, they would exit the automotive market as quickly as they came.

In reality, semiconductor manufacturers have been diligent to demonstrate their commitment to the automotive industry, investing in automotive-grade manufacturing capabilities, emphasizing functional safety and cybersecurity, and positioning themselves not only as component suppliers, but as trusted innovation partners that serve as the focal point for the industry’s transition toward software-defined vehicles.

Nevertheless, Intel’s sudden departure from the automotive market is likely to stir anxieties. However, there are steps that automakers can take to minimize their risk exposure that semiconductor manufacturers can support to better establish their position as the automaker’s foremost innovation partner.    

The key to reducing the risk exposure associated with compute hardware is to break the link between specific hardware platforms or suppliers and the value-creating, differentiating software applications that OEMs layer on top. Middleware provides a vital abstraction layer that can embolden automakers to choose the latest and greatest in semiconductor technologies, with the confidence that it will support any software engineering investment made to date, and that the OEMs’ continued software development with the hardware platform in question will retain its value if the automaker needs to pivot to an alternative.

Intel’s departure from the automotive market is a reminder of the danger to automakers in entrusting too much to a single partner, but automakers should be cautious about overreacting and missing out on the best that the innovation engines of the consumer electronics space have to offer. Middleware abstraction layers, and the disentangling of hardware and software development, more generally, holds the key to allowing automakers to taking full advantage of what semiconductor giants have to offer, without taking their hand off the tiller, and remaining robust to changes taking place outside of their captive supply chain.

Finally, semiconductor suppliers themselves should engage enthusiastically with initiatives such as SOAFEE that are developing the necessary abstraction layers and cloud-based approaches to disentangle hardware and software development. In the first case, any strategy built around vendor lock-in is bound to fail before it begins, and while there is a risk that OEMs will take advantage of hardware abstraction to switch away from a supplier, it can prove equally helpful in attracting an OEM away from a competitor and to one’s own hardware platform. Moreover, it can ease the transition from aging platforms to the next-generation platforms, keeping automakers in lock-step with the consumer electronics industries they crave to emulate.