SK Hynix Launches Next-Generation HBM4 for NVIDIA’s Rubin AI GPUs, Solidifying Dominance in the HBM Market Over Micron and Samsung

SK Hynix Launches Next-Generation HBM4 for NVIDIA’s Rubin AI GPUs, Solidifying Dominance in the HBM Market Over Micron and Samsung

SK hynix has once again asserted its dominance in the High Bandwidth Memory (HBM) sector by becoming the inaugural supplier of HBM4 modules to NVIDIA for the testing of its Rubin project samples.

SK Hynix Leads the Charge for NVIDIA’s HBM4 Orders, Outpacing Micron and Samsung

Understanding the significance of HBM4 is pivotal for grasping future trends in computational technology. This cutting-edge memory standard uniquely integrates memory and logic components into one cohesive package, marking a groundbreaking advancement in the industry.

The competition for HBM4 supremacy has been a focal point among major South Korean tech giants. While Samsung has made strides with its HBM4 offerings, specific yield rates remain undisclosed. Nevertheless, Samsung recently garnered positive attention with the adoption of its HBM3E module by AMD, boosting its reputation potentially enough to position it favorably as a candidate in NVIDIA’s “pool of suppliers”for HBM4.

NVIDIA Expected To Mass Product Next-Gen R100

It’s also worth noting that NVIDIA is planning a “multi-sourcing”strategy similar to what it employed with HBM3, which could see it collaborating with both Micron and potentially Samsung for its HBM4 needs.

With the introduction of HBM4 modules, NVIDIA’s Rubin GPU architecture is set to launch around Q4 2025, with customer qualification expected to commence in September of that year. This timing highlights the urgency and competitive dynamics in the race for advanced memory solutions.

Source & Images

Leave a Reply

Your email address will not be published. Required fields are marked *