SK Hynix has the graphic DRAM fastest according to own statements presented . The 8 GB working after large blocks of type graphics DDR6 (GDDR6) company information high end graphic cards with 384-bit I/o can process up to 768 GB graphics data with an I/O data rate per PIN by 16 Gbit / s per second.
Specifies the bandwidth per pin of of GDDR5 modules SK Hynix with 8 Gbit / s. As a result a maximum transfer rate of 32 Gbytes per second currently. The operating voltage to 10 percent to decline despite the double speed per PIN.
“SK Hynix will respond with the performance of the industry’s fastest GDDR6 store proactively on high-quality and high-performance graphics memory solutions”, Jonghoon said Vice President and head of the SK Hynix DRAM product development Oh. “We will help our customers to improve the performance of their high end graphics cards.”
GDDR6 is required for growth areas such as artificial intelligence, virtual reality, self-propelled vehicles and displays with a resolution of more than 4 K according to the company. SK Hynix also anticipates an increasing demand to GDDR memory modules for graphics cards. Gartner guess that the average density of DRAM memory cards of 2.2 GB go on into this year 4.1 GB in 2021. This corresponds to an average annual growth rate 17 per cent. The new chips in a 20-nanometer process are produced
. Mass production is to start in time for a launch of the first high end graphic cards with GDDR6 memory in the spring of 2018. Within this time frame, alsoSamsung and Micron.
a further new development had announced earlier this month the company. With the world’s first TLC-3D-NAND-flash-chip, which consists of 72 layers SK Hynix rose 32 GByte capacity per chip. Also here, the South Korean party overtook its competitors Samsung and Toshiba, which are made currently 64-layer modules and/or switch the production soon to 64 layers.
get to know the building blocks of enterprise cloud platform in this audio Webinar. Learn how to achieve maximum freedom and flexibility for your applications. More outcome achieved with less input – specific application examples.
[withmaterialfromChoMu-Hyun ZDNet.com ]