SK hynix announced today that it has successfully developed what it is calling the world’s best-performing HBM3E next-generation DRAM for AI applications. The new memory can process data up to 1.15 terabytes a second, which is equivalent to “processing more than 230 Full-HD movies of 5GB-size each in a second,” according to the company.
“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.”
The company said that the successful development of HBM3E, the extended version of HBM3 which delivers the world’s best specifications, comes on top of its experience as the industry’s sole mass provider of HBM3. With its experience as the supplier of the industry’s largest volume of HBM products and the mass-production readiness level, SK hynix plans to mass produce HBM3E from the first half of next year and solidify its unrivaled leadership in AI memory market.
According to the company, the latest product not only meets the industry’s highest standards of speed, the key specification for AI memory products, but all categories including capacity, heat dissipation and user-friendliness.
Discussion (1 reply)
Join Discussion →That's the throughput of the ram but in what configuratrion? What bus can support that? What CPU's can consume that. Not saying it's a bad thing but there is a lot of guess work around that statement... lets see if the article exposes more...
Nope doesn't seem to expand on the actual channels/config. But it's targeted to AI specific tasks which means it'll have dozens if not hundreds of CPU's making memory requests as long as the BUS speed is there.