SK Hynix, a top company in South Korea’s semiconductor industry, has achieved impressive progress in memory technology through its new invention, called HBM3E. This big step forward represents the fifth stage of High Bandwidth Memory (HBM) technology, building on the success of its forerunners – HBM, HBM2, HBM2E, and HBM3.
HBM3E shows SK Hynix’s skills, boosting data speed hugely using its strong HBM3 memory know-how. This achievement cements SK Hynix’s top spot and readies mass production for early next year. At the core of HBM3E’s amazing performance is its unmatched capacity, reaching 1.15 terabytes (TB) per second. To grasp this, think about handling 230 Full HD movies, each 5 GB, in just one second.
One of the key innovations that set the HBM3E apart is the use of a cutting-edge technology known as Advanced MR-MUF (Mass Reflow Molded Underfill). The latter can make the cooling process better by 10% than the one before it. As a result, the memory will work faster and smoother. The new memory stacks feature a data transfer rate of 9 GT/s, which exceeds the company’s HBM3 stacks by a whopping 40%. The best thing about HBM3E is its incredible power – processing 1.15TB of data per second. Imagine 230 Full-HD movies (5GB each) in just one second.
Gizchina News of the week
Plus, the HBM3E is compatible with models that are already on the market. So users do not need to make many changes when switching between their old and new memories. In other words, if you have a device that runs on a memory of the previous generation, there is no need to make big changes. The HBM3E memory will work with all known models.
HBM3E: The Memory Marvel that’ll ‘Blow’ Your Mind!
This big step has caught the eye of giants like NVIDIA. Ian Buck, NVIDIA’s VP, is thrilled to keep collaborating on HBM3E for the next AI computing phase. In a market needing faster data handling, SK Hynix’s HBM3E is a game-changer. It’s set to reshape AI memory technology. As it hits the market next year, its power to boost AI computing and speed up technology progress shines through.
“We have a long history of working with SK hynix on High Bandwidth Memory for leading-edge accelerated computing solutions,” said Ian Buck, Vice President of Hyperscale and HPC Computing at NVIDIA. “We look forward to continuing our collaboration with HBM3E to deliver the next generation of AI computing.”
SK Hynix shines as a top “mass producer” of HBM3 technology. It’s set to meet rising AI storage needs, boosting HBM3E production in early 2024. Their solid promise to enhance the AI memory market through top-notch HBM items assures a swift, game-changing business shift.
As AI and ML are becoming popular on a daily basis, the need for faster data processing becomes more urgent than ever. SK Hynix’s new HBM3E technology, with its unseen speed, is emerging as an innovation leader with the potential to transform the field of AI computing. When this tech arrives, it will reshape AI memory and fuel big changes.
