SK Hynix Begins Mass-Production of HBM2E High-Speed DRAM
SK hynix announced that it has started the full-scale mass-production of high-speed DRAM, 'HBM2E', only ten months after the Company announced the development of the new product in August last year.
SK hynix's HBM2E supports over 460 GB (Gigabyte) per second with 1,024 I/Os (Inputs/Outputs) based on the 3.6 Gbps (gigabits-per-second) speed performance per pin. It is the fastest DRAM solution in the industry, being able to transmit 124 FHD (full-HD) movies (3.7 GB each) per second. The density is 16 GB by vertically stacking eight 16 Gb chips through TSV (Through Silicon Via) technology, and it is more than doubled from the previous generation (HBM2).
HBM2E boasts high-speed, high-capacity, and low-power characteristics; it is an optimal memory solution for the next-generation AI (Artificial Intelligence) systems including Deep Learning Accelerator and High-Performance Computing, which all require high-level computing performance. Furthermore, it is expected to be applied to the Exascale supercomputer - a high-performance computing system which can perform calculations a quintillion times per second - that will lead the research of next-generation basic and applied science, such as climate changes, bio-medics, and space exploration.
"SK hynix has been in the forefront of technology innovation that contributes to human civilization with achievements including the world's first development of HBM products," said Jonghoon Oh, Executive Vice President and Chief Marketing Officer (CMO) at SK hynix. "With the full-scale mass-production of HBM2E, we will continue to strengthen our presence in the premium memory market and lead the fourth industrial revolution."Annotation
■ HBM (High Bandwidth Memory)
- High performance, high bandwidth memory products that adopt TSV technology to dramatically accelerate data processing speed over traditional DRAMs.
■ TSV (Through Silicon Via)
- An interconnecting technology that connects the upper and lower chips through thousands of fine holes on DRAM chip.
- Delivers data, commands, and currents through column-shaped paths that penetrate the entire silicon wafer thickness after stacking multiple DRAM chips on the buffer chip.
- Up to 30% decrease in size and up to 50% decrease in power consumption over existing packaging methods.
■ Standards for data process speed conversion
- 1 GB = 8Gb
- 3.6 Gbps per pin with 1024 data I/Os (Inputs/Outputs) = 3686.4 Gbps
- 3686.4 Gbps / 8 = 460.8 GB/s (Gb -> GB conversion)
SK Hynix Releases PE8000 NVMe PCIe Gen4 SSD - PCIe 4.0 up to 6,500 MB/sec - 04/08/2020 09:05 AM
SK hynix announced the availability of the “PE8000” series, its newest and most advanced line-up of enterprise Solid State Drives (eSSDs). ...
SK Hynix to have Superfast 4800 Mbps DDR5 memory on its roadmap - 04/02/2020 06:42 PM
Starting next year we'll likely see the first commercial DDR5 implementations, often shortly followed by the consumer segment. SK Hynix shares a few words on that as they are preparing DDR5 that supp...
SK Hynix Likely to announce first 128-layer NAND P31 M.2 SSDs next week already - 12/30/2019 11:00 AM
Much like an onion, the more layers the more mass and volume. In this case, NAND mass storage. A while ago SK Hynix already shared word that they delivered the first samples of its solution products ...
SK Hynix Outs Engineering Samples of Terabyte-Level Solutions Based on a 128-Layer 4D NAND - 11/20/2019 09:29 AM
The era of Terabyte (TB) memory is arriving with the innovative development of NAND Flash technology. SK hynix delivered thefirst samples of its solution products based on its 128-layer 4D NAND (c...
SK hynix Develops 1Znm 16Gb DDR4 DRAM - 10/21/2019 08:51 AM
SK Hynix has developed 1Znm 16Gb (Gigabits) DDR4 (Double Data Rate 4) DRAM. As 16Gb is the industry's largest density for a single chip, the total memory capacity per wafer is also the largest of the...
Senior Member
Posts: 1991
Joined: 2013-06-04
On affordable GPUs they wouldn't be needed, or the difference would be negligible. They're usually paired with the right GPU for the best combined performance.
Everything counts for the finally price: GPU, VRAM, the bus lanes, etc.
Senior Member
Posts: 1259
Joined: 2013-02-22
I think these could be viable and possibly even cheaper than standard gddr ram if they didn't need to use the expensive interposer with its reduced yields on top. I know its not so simple but picture if u could have the gpu with a single stack beside it instead of having 10 or however many chips surrounding the gpu.
Senior Member
Posts: 3105
Joined: 2016-08-01
The more they produce the better if we have any chance seeing hbm costing about as much as gddr we need more production and better yields!
Senior Member
Posts: 157
Joined: 2016-12-31
What speed, too bad they're expensive to use on affordable graphics.