Samsung Develops 8-Layer HBM4E to Meet NVIDIA's Requirements

Samsung Electronics is developing an 8-layer HBM4E product to meet NVIDIA's requirements, intended for custom high-bandwidth memory NVHBM. This solution reduces the number of stacked layers compared to the originally planned 12-layer and 16-layer products. NVIDIA's target speed is 17 to 18 Gbps, which is about a 20% increase from the earlier sample speed of 14.4 Gbps. The 8-layer design can reduce the challenges of wafer thinning, precision stacking, and backend packaging, alleviating yield pressure and enhancing supply capability. NVHBM is expected to be applied starting with the Rubin Ultra AI GPU, which will launch next year. The GPU interconnect scale of Rubin Ultra will expand from the current 72 chips to a maximum of 576 chips, allowing more high-speed GPUs to work together to compensate for the decline in single GPU memory capacity. Samsung has the design and production capabilities for DRAM, logic chips, and basic bare chips, providing a one-stop service for custom HBM. Industry insiders believe that Samsung has demonstrated high transmission speeds with HBM4, and as NVIDIA shifts its competitive focus to high-speed transmission, its related capabilities may receive more attention.

This content is provided for general informational purposes only and doesn't constitute financial, investment, legal, or tax advice. Any events, rewards, online promotions, or related information mentioned herein should not be considered a recommendation, solicitation, or invitation to purchase, sell, trade, or otherwise deal in any crypto assets. Crypto assets are highly volatile and may result in loss. The availability of WEEX services, products, and related events may vary by region. You are responsible for ensuring that your participation is in accordance with applicable local laws and regulations.