Samsung to launch world’s first 6th-gen HBM4 memory for Nvidia AI chips

Samsung Electronics is preparing to commence mass production of its sixth-generation high-bandwidth memory, known as HBM4, as early as next week, according to industry reports. The new memory chips are designed for use in next-generation graphics processing units, particularly those produced by Nvidia for advanced artificial intelligence systems.

Production Timed with Nvidia’s AI Roadmap

South Korea’s Yonhap News Agency reported Sunday that Samsung’s production schedule has been aligned with Nvidia’s plans to launch its forthcoming AI accelerator, codenamed Vera Rubin. Shipments of the HBM4 chips are expected to begin after the Lunar New Year holiday. Samsung has already passed Nvidia’s stringent quality certification process and secured purchase orders, confirming the chips' readiness for integration into high-performance computing platforms.

Next-Generation Technology for AI

The HBM4 memory represents a significant step beyond the current industry-standard fifth-generation HBM3E chips, offering enhanced bandwidth and efficiency critical for training and running large generative AI models. As the demand for AI computing power grows, HBM4 is anticipated to become a foundational technology in data centers and advanced workstations. Nvidia’s adoption of the memory in its Vera Rubin platform underscores its strategic importance.

Market Position and Industry Impact

Samsung has reportedly increased the volume of HBM4 samples for customer-side module testing, indicating robust preparatory steps ahead of full-scale manufacturing. The move solidifies Samsung’s competitive stance in the global HBM market, where it vies with rivals such as SK Hynix. The company’s ability to deliver cutting-edge memory solutions is crucial as the semiconductor industry increasingly pivots toward AI-driven hardware development.