The successful rollout highlights SEMIFIVE’s turnkey capabilities, managing the entire lifecycle from front-end design to final packaging and testing. Handling a die of this magnitude requires precise control over power consumption and heat dissipation, proving the firm’s technical maturity in advanced-node manufacturing. This milestone follows a string of successful production runs, including AI chips for Hanwha Vision’s security systems and high-performance computing units for Japanese clients.
SEMIFIVE Launches HyperAccel AI Chip Production on Samsung 4nm Node
Seoul-based custom semiconductor provider SEMIFIVE has initiated mass production for HyperAccel’s Bertha, a data center AI inference accelerator. The project marks the company’s first large-scale deployment on Samsung Foundry’s 4nm process, utilizing a complex “Big Die” architecture exceeding 500 mm² to support high-performance language model workloads.
Financial indicators suggest a significant shift in the company’s business model toward recurring revenue. SEMIFIVE reported KRW 42.3 billion in new mass-production orders during the first half of 2026, nearly doubling the total intake from the previous year. Quarterly growth remains robust, with a 71% increase between the first and second quarters. As global demand for specialized inference hardware climbs, the company is positioning itself to capture long-term supply contracts, moving beyond one-time design projects to establish a stable, high-value manufacturing pipeline.
Comments (0)
No comments yet. Be the first!