What if the problem isn't EVs, but architecture?

Much of the discussion around automotive transformation focuses on electrification, tariffs or Chinese competition. But GM appears to believe the industry's biggest challenge sits much deeper in the vehicle itself.

Its diagnosis is that automakers are attempting to deliver software-defined experiences, AI-powered features and continuous updates on top of electrical architectures that were never designed for that level of complexity. In GM's view, the bottleneck isn't the product portfolio, it's the technical foundation beneath it.

That belief is driving one of the most ambitious architecture programmes in the Western automotive industry.

At its GM Forward event in late 2025, the company unveiled plans for a next-generation centralized compute platform scheduled to launch in 2028. Unlike the zonal architectures many OEMs are now adopting, GM is aiming for a more centralized approach built around high-speed Ethernet networking and Nvidia Drive AGX compute.

The distinction is important.

Most of today's vehicles still distribute intelligence across dozens, sometimes hundreds, of electronic control units (ECUs). Even many next-generation architectures remain partially decentralized. GM's strategy is to consolidate vehicle intelligence into a smaller number of powerful computing domains, creating a platform capable of supporting future software, AI, and automated driving functions more efficiently.

The company is already taking steps in that direction. In April 2026, GM deployed Google Gemini across eligible Cadillac, Chevrolet, Buick, and GMC vehicles, making it the only Detroit automaker currently operating a generative AI assistant at significant production scale.

Taken together, these moves suggest GM sees software capability not as a feature layer but as an architectural challenge.

The strategy also benefits from something many competitors lack, financial flexibility. GM entered 2026 with the strongest balance sheet of Detroit's Big Three, giving it greater freedom to invest in long-cycle technology programmes while continuing to return capital to shareholders.

However, there is an important caveat.

GM may be leading Detroit on vehicle architecture, but it is not leading globally. Tesla has operated a highly centralized computing model for years, while Chinese OEMs such as NIO, Xpeng, and Li Auto already have comparable architectures in production vehicles today.

Meanwhile Rivian has taken a vertically integrated, software-first approach with a relatively centralized architecture that enables rapid OTA iteration and tight hardware-software co-development.

That means GM's bet is not about establishing a new frontier. It is about closing a gap before it becomes strategically unmanageable. Whether that gap can be closed will depend on execution. Delivering a centralized architecture across both EV and ICE portfolios by 2028 is a significant engineering challenge. It will also require enough computing headroom to accommodate future AI workloads, an area where even some industry leaders have encountered limitations.

For suppliers, the implications are substantial. Centralized architectures typically shift value away from distributed hardware and toward silicon, software, networking and systems integration. As vehicles become more compute-centric, sourcing decisions become concentrated into fewer, larger and more strategic technology partnerships.

The broader question is whether architecture becomes the primary competitive battleground of the next decade. If GM is correct, the companies that win may not be those with the best products today, but those with the most capable software foundations tomorrow.

This is just one of three very different strategic bets currently being made by Detroit's Big Three. Ford is pursuing a product-and-cost-led approach, while Stellantis is prioritizing balance-sheet repair and portfolio focus. Together, they form an unintended industry experiment on how best to respond to the pressures reshaping automotive markets. The full article explores all three approaches, what they mean for suppliers and investors, and which diagnosis the market currently appears to believe. Read the full article here: Detroit's big three are placing different bets. Will all approaches succeed? James Alford, Consulting Manager at SBD Automotive

How SBD can help

SBD Automotive can help benchmark your position against the wider industry and identify where action is needed most. To explore how these trends impact your strategy, architecture and supplier roadmap, get in touch with SBD Automotive for a deeper discussion. Email info@sbdautomotive.com