I have spent much of my career at the point at which processors, software, and automotive systems converge. Today, that intersection matters more than ever.

The automotive industry is undergoing one of the most significant technology transformations in its history, and I have seen firsthand how quickly the role of automotive compute is expanding.

With the long-standing vision of the software-defined vehicle (SDV) increasingly a reality on our roads, vendors are naturally keen to understand what pervasive AI brings to the vehicle – from agentic AI that interprets human intent and acts on the in-cabin experience to physical AI that perceives and adapts vehicle behavior within the physical world.

Running these advanced software workloads requires a high-performance compute foundation. But as its importance grows, control of that foundation becomes an increasingly strategic consideration.

OEMs need long-term economic control – the ability to choose suppliers, maintain negotiating leverage, reuse software across vehicle programs, and evolve hardware platforms without vendor lock-in. The instruction set architecture (ISA) they choose is strategically critical because vehicle platforms stay in production for over 10 years and on the road much longer, meaning architectural decisions influence development costs, supplier relationships, and technology choices across multiple generations.

For automotive OEMs and Tier 1s, this should be the fundamental question. The choice of architectural foundation should be driven by more than processor performance; it should prioritize long-term economic control.

Vehicle platforms are measured in decades, not years. A platform might stay in production for more than 10 years and remain on the road for potentially far longer than that.

Software developed for that platform must be maintained until the last car goes for scrap. Carmakers also have a duty to ensure that the vehicle retains the same or better functional safety and cybersecurity capabilities that it did when it was first sold.

The bottom line: any architectural decision made today will influence development costs, supplier relationships, software investments, and technology choices across multiple generations.

In my view, automotive OEMs and Tier 1s need to think about compute architecture strategy in the same way they think about other strategic technology decisions: as a source of long-term economic control and competitive advantage.

That is why I believe RISC-V represents such an important opportunity for the automotive industry.

You may have heard me talk about automotive economic control, but I don’t mean that automakers need to own every piece of technology. I mean they need choice.

OEMs should have the ability to choose their suppliers and implementation partners, maintain negotiating leverage, and reduce concentration risk. They should also be able to reuse software investments across vehicle programs and generations.

Most importantly, they should be able to evolve their hardware platforms without being locked into a particular vendor or technology roadmap they don’t control.

This is where the instruction set architecture (ISA) you base that hardware infrastructure on becomes strategically important.

An open architecture like RISC-V creates a common foundation that multiple companies can build upon. It doesn’t eliminate competition: it can actually expand it.

Semiconductor vendors can differentiate their implementations. IP providers can innovate around the architecture. Software companies can build broader ecosystems. And OEMs can evaluate different solutions while maintaining a degree of architectural consistency.

At MIPS, we firmly believe that standardization and differentiation do not have to be opposites. In fact, I believe the best architectures give the industry both.

RISC-V is an open standard ISA, but what makes it particularly compelling to me is what happens above that base architecture.

Companies can differentiate their processor implementations, add capabilities, optimize for specific workloads, and build complete systems around the architecture without having to create a proprietary ISA from scratch.

That creates a fundamentally different model for automotive computing.

A common architecture can provide consistency while the implementation remains highly differentiated. An OEM can pursue solutions optimized for real-time control, functional safety, AI, or other vehicle-specific workloads. Semiconductor companies can develop differentiated processors. And the broader ecosystem can innovate without every participant having to control the underlying ISA.

One of the most powerful aspects of RISC-V is that it creates a common foundation without dictating what companies build on top of it.

And because RISC-V is an open industry standard governed by a global ecosystem of industry and academic stakeholders rather than any one company, it offers the potential for greater long-term architectural stability.

That matters enormously when your product lifecycle is measured in vehicle generations rather than product refresh cycles.

Subscribe to RISC-V Automotive Insights for exclusive perspectives from companies shaping tomorrow’s software‑defined vehicles.

Delivered quarterly to your inbox. No detours.

We should think beyond individual processors and start thinking about the vehicle as a computing platform.

Today, a vehicle can contain computing systems spanning body and comfort functions, zone and domain control, gateways, real-time and safety-critical applications, ADAS, centralized compute, cockpit systems, and increasingly sophisticated, pervasive AI workloads.

Each of these requires workload-specific silicon designed around a particular set of tasks, rather than identical, one-size-fits-all processors. But there is real value in greater architectural consistency across the SDV.

RISC-V enables the automotive industry to carry a common ISA across a much broader portion of the vehicle, while allowing processors to remain differentiated for their specific requirements. For software teams, this creates greater opportunities to reuse code, tools, and expertise, translating into real economic benefits.

Software development is becoming one of the largest components of overall vehicle engineering. If the underlying compute architecture changes from one processor class to another, software teams are forced to re-engineer, validate, and maintain different environments, use different tools, and employ different methodologies.

A common architecture can reduce that fragmentation, creating more opportunities to reuse software, development tools, and engineering expertise across multiple processor classes and vehicle programs. The result is less duplication, and a greater return on software investments.

For an industry investing billions of dollars in SDVs, that is not a small benefit.

The proliferation of AI makes this architectural question even more important.

We are moving toward a world where AI will be embedded throughout the vehicle, in perception and sensing, driver assistance, safety systems, predictive maintenance, cockpit experiences, and many other applications.

But AI workloads are not static.

As innovation and technology evolve, models will change. Accelerators will change. Memory architectures will change. New workloads will emerge that we cannot fully anticipate today.

So the question isn’t simply whether a processor can run today’s AI workload. It is whether the compute architecture gives an automaker the freedom to adapt, to build in headroom for the future workloads, and retain software portability when hardware needs to evolve to compute the unknown workloads of the future.

This is where RISC-V can provide an important advantage. It allows the underlying software foundation to remain consistent while implementations adapt to new workloads and technologies.

Today, there is a clear need in the automotive industry for platforms that combine openness with the level of performance, scalability, efficiency, and differentiation required by demanding applications.

RISC-V provides the open architectural foundation. Our job at MIPS is to build on it with the processor and platform technologies that demanding real-world applications require.

Through our world-class engineering and design team, we bring decades of processor experience to RISC-V, with a focus on differentiated processor IP, software, and the technologies needed to support increasingly complex workloads.

We also believe that openness should not mean commoditization. This is a distinction that I believe is particularly important in automotive.

An OEM should be able to establish a common architectural foundation while still differentiating its vehicle. A semiconductor company should be able to differentiate its implementation. And the software ecosystem should be able to build reusable technologies that span multiple systems.

That is the model we are helping advance at MIPS, and it’s enabled by the open, modular nature of RISC-V. RISC-V gives the automotive industry both standardization and the freedom to create something extraordinarily powerful.

And perhaps most importantly, RISC-V gives the companies building the vehicle greater influence over the technology that will define its future, via an ecosystem that enables and encourages innovation, differentiation, and long-term economic control. For companies ready to take advantage of that freedom, MIPS provides an easy way to get started.