Samsung, SK Hynix To Provide High Bandwidth Memory for Nvidia's Pascal GPU
A report from Korea says that Nvidia will get second generation High Bandwidth Memory (HBM) for its upcoming Pascal GPU by both SK hynix and Samsung Electronics. Production will start in Q1 2016 following the pilot production and reliability tests expected to complete later this year, Business Korea reports.
SK Hynix has been already supplying first-gen HBM to AMD and Nvidia ahead of Samsung. Micron uses Hybrid Memory Cubes (HMC), a technology similar to HBM.
In addition to the HBM2 bandwidth boost, Nvidia says it will enhance the performance of Pascal GPUs with the NVLink interconnect technology, which is said to allow data sharing at rates 5 to 12 times faster than the traditional PCIe Gen3 interconnect.
Nvidia expects that HBM equipped graphics cards will be 'mainstream' by 2018.
HBM is a high-performance RAM interface that stacks up DRAM modules in four layers. Using Through Silicon Via (TSV) technology, the stacked memory chips can process data 4 to 8 times faster than conventional DDR memory, and consume 40 percent less power. TSV technology is a technique that enables a vertical electrical connection of chips by making tiny holes in chips, rather than connecting chips using wires. After DRAM chips are sliced thinner than half of the thickness of a piece of paper, they are pierced to contain hundreds of fine holes, and vertically connected through electrodes that pass through the holes of the chips.
The second-gen HBM chips process data twice as fast as the first-gen HBM modules, which were unveiled by SK Hynix in late 2013. The storage capacity of the 8 Gb HBM chips is said to have increased fourfold.