onsemi (NASDAQ: ON) used its 2026 Investor Day on September 16 to lay out a long-term growth strategy centered on a $213 billion addressable market it expects to materialize by 2030, spanning automotive, industrial, AI data centers and what the company calls "emerging applications." President and CEO Hassane El-Khoury framed the opportunity around a specific engineering bottleneck: power density, or the ability to deliver more performance from less energy and less physical space. "Power has become a foundational technology layer that will help shape the pace of innovation, economic growth and global competitiveness," El-Khoury said.
onsemi is building that pitch on three technology platforms. The first, its high-voltage portfolio, spans silicon, silicon carbide and gallium nitride, and now includes vertical gallium nitride (vGaN), a process the company says is exclusive to onsemi and aimed at applications where conventional power technologies run out of headroom. The second, called Treo, combines high-precision analog, high-density digital and high-voltage circuitry to let systems monitor operating conditions and respond to load changes in real time. For engineers specifying components in inspection or motion-control systems, that combination of analog precision and high-voltage switching in one platform is the more consequential detail than the market-size figure.
The third platform, Embedded Power Platform (EPP), uses the silicon wafer itself as the package rather than a passive housing, letting onsemi embed FETs, drivers and controllers together and co-optimize them for electrical, thermal and mechanical performance. The company says that approach raises power density and shortens development time compared with assembling discrete components separately.
onsemi is also betting that AI investment moves beyond training infrastructure and into machines that sense and act in physical environments, a shift it calls "physical AI." It describes four capabilities such systems need: power, sensing, control and "connected compute" to process and share data at the edge. El-Khoury said the company's ability to combine those functions through system-level co-design is what distinguishes its approach from competitors chasing the same opportunity. The company points to robots, vehicles, factories, medical systems and other autonomous platforms as the places that convergence will show up first.
