Compal Electronics is heading to San Jose with a pitch aimed squarely at the people building the world’s largest AI systems. The company plans to show AI factory infrastructure designed specifically for Nvidia accelerated computing platforms at the OCP Global Summit 2026.
The headline feature is an 800VDC power architecture paired with megawatt-class infrastructure. The goal is to help AI deployments cope with power density demands that keep climbing.
What Compal is bringing to San Jose
The summit runs October 12–15, 2026, at the San Jose McEnery Convention Center. Compal will be set up at Booth #E65.
Its showcase covers four connected areas: power delivery, cooling, high-density compute and rack-scale engineering. Rather than selling a single box, Compal is presenting these as integrated systems that work together.
The term “AI factory” is the industry’s shorthand for facilities built to produce AI outputs at industrial scale. Compal says its solutions are meant to help these operations grow smoothly, starting at the platform level and stretching all the way to megawatt deployments.
The 800VDC piece deserves a closer look. VDC stands for volts of direct current, and 800VDC refers to the voltage level Compal’s power architecture is built around. The stated purpose is to handle rack power densities that are rising as AI deployments get larger and more demanding.
A crowded summit with one dominant topic
Compal will not be lonely in San Jose. The summit sold out, drawing approximately 15,000 attendees, with many more turned away at the door.
The agenda leaves little doubt about what is driving that demand. AI appeared prominently in over ninety percent of the event’s press releases this year.
Nvidia is also playing a direct role at the event. The chipmaker contributed open designs for its Vera Rubin rack-scale architecture.
That contribution helps explain why hardware partners like Compal are tailoring their offerings so closely to Nvidia platforms. When the dominant chip supplier publishes open rack-scale designs, the companies building power, cooling and compute systems have a clearer blueprint to work from.
Why power and cooling are the new bottleneck
Compal’s framing reflects a shift toward rising power density, meaning more energy concentrated into each rack as AI systems scale up. By bundling power delivery, cooling, compute and rack-scale engineering into integrated systems, Compal is positioning these as co-designed solutions optimized for Nvidia platforms.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.