Micron start volume producting of HBM3E to meet demand in AI server market

Micron Technology commenced volume production of its HBM3E (High Bandwidth Memory 3E) solution to address exponentially growing demand in AI server market. Partnering with NVIDIA, Micron's 24GB 8H HBM3E will power the upcoming NVIDIA H200 Tensor Core GPUs, ushering in a new era of artificial intelligence (AI) solutions. This development underscores Micron's leadership in providing memory solutions tailored to the demanding requirements of AI applications. Superior Performance: Micron's HBM3E offers a pin speed exceeding 9.2 gigabits per second (Gb/s), translating to over 1.2 terabytes per second (TB/s) of memory bandwidth. This lightning-fast data access is ideal for AI accelerators, supercomputers, and data centers, enabling rapid processing of massive datasets. Exceptional Efficiency: Micron's HBM3E boasts approximately 30% lower power consumption compared to competing solutions. This enhanced energy efficiency is crucial for AI workloads, reducing operational expenses and supporting the growing demand for AI applications. Seamless Scalability: With a capacity of 24GB, Micron's HBM3E provides data centers with seamless scalability for AI applications. Whether for training complex neural networks or accelerating inferencing tasks, Micron's solution delivers the required memory bandwidth to meet evolving AI demands. Sumit Sadana, Executive Vice President and Chief B...