BOISE, Idaho - Micron Technology, Inc. (NASDAQ:MU), a prominent memory and storage solutions provider, has commenced volume production of its High Bandwidth (NASDAQ:BAND) Memory 3E (HBM3E) solution, the company announced today. Micron's 24GB 8H HBM3E is set to be integrated into NVIDIA (NASDAQ:NVDA) H200 Tensor Core GPUs, which are expected to ship in the second quarter of 2024.
The new HBM3E solution by Micron boasts a superior performance with a pin speed exceeding 9.2 gigabits per second, delivering over 1.2 terabytes per second of memory bandwidth. This advancement is aimed at meeting the rapidly growing demands of artificial intelligence (AI) applications, supercomputers, and data centers by providing swift data access for AI accelerators.
Micron's HBM3E stands out for its exceptional energy efficiency, consuming approximately 30% less power than its competitors' HBM3E offerings. This feature is particularly significant as it can contribute to reducing data center operating costs, a key consideration for businesses relying on large-scale AI operations.
The company's HBM3E also offers seamless scalability with its 24 GB capacity, facilitating the expansion of AI applications ranging from training extensive neural networks to accelerating inferencing tasks.
Sumit Sadana, executive vice president and chief business officer at Micron, highlighted the importance of memory bandwidth and capacity for AI workloads and expressed confidence in Micron's position to support the anticipated AI growth.
In addition to the current production, Micron is advancing its leadership by sampling a 36GB 12-High HBM3E variant in March 2024, which is expected to deliver similar performance and energy efficiency advantages.
Micron's development of the HBM3E design utilizes its 1-beta technology and advanced through-silicon via (TSV), among other innovations, contributing to a differentiated packaging solution. The company is also part of TSMC's 3DFabric Alliance, playing a role in the future of semiconductor and system innovations.
The announcement aligns with Micron's sponsorship of the NVIDIA GTC, a global AI conference starting on March 18, where the company plans to elaborate on its AI memory portfolio and roadmaps.
This news is based on a press release statement from Micron Technology, Inc.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.