Rebellions Inc., a leading AI semiconductor company based in South Korea, today announced a collaboration with Marvell Technology, Inc. (NASDAQ: MRVL) to offer high-performance, energy-efficient AI systems tailored for regionally driven and sovereign-backed AI initiatives across APAC and the Middle East.
As AI infrastructure becomes increasingly central to national competitiveness, a shift is underway from standardized, GPU-based architectures toward domain-specific systems built around custom ASICs. This change is being driven not only by global hyperscalers, but also by sovereign-backed initiatives and regional cloud providers seeking infrastructure that delivers scalability, efficiency, and control.
This collaboration will enable Rebellions to design customer-specific AI accelerators utilizing Marvell® custom platforms—leveraging advanced packaging, high-speed SerDes and die-to-die interconnects resulting in a tightly integrated, end-to-end rack solution built for high-performance, energy-efficient AI inference at scale.
"AI infrastructure is undergoing a fundamental shift—organizations are no longer settling for one-size-fits-all solutions," said Sunghyun Park, CEO of Rebellions. "By collaborating with Marvell, we're combining deep expertise in AI chip design with cutting-edge silicon integration to deliver custom AI infrastructure designed to meet the real-world needs of sovereign entities."
"Custom AI infrastructure is key to unlocking the next wave of data center innovation," said Will Chu, Senior Vice President and General Manager of Custom Cloud Solutions at Marvell. "We look forward to collaborating with Rebellions to offer next-generation custom AI infrastructure tailored for performance, efficiency and scalability."
