Samsung Electronics has begun mass producing the industry's first two gigabyte (GB), low power double-data-rate 3 (LPDDR3) memory, using 30 nanometer class technology, for next-generation mobile devices.
Samsung started mass production of the industry's most advanced mobile DRAM (dynamic random access memory) chip only ten months after it began producing the industry's first 30nm-class based 2GB LPDDR2 memory in October, 2011. The new LPDDR3, which marks the first time a 2GB LPDDR3 density is available in one space-saving package, utilizes four LPDDR3 chips stacked together. LPDDR3 is needed for fast processors, high resolution displays and 3D graphics in tablets and smartphones.
"We will embrace greater technical cooperation with industry leading mobile device makers, as we continue to provide timely next-generation memory solutions like 2GB LPDDR3 DRAM, in helping to accelerate growth of the mobile memory market," said Wanhoon Hong, executive vice president, Memory Sales & Marketing, Samsung Electronics. "Furthermore, with this rapid introduction of 2GB LPDDR3, we are moving very assertively to expand our global leadership on the premium side of our extensive memory portfolio."
The new 2GB LPDDR3 DRAM can transfer data at up to 1600 megabits per second (Mbps) per pin, which is approximately 50 percent faster than a LPDDR2 DRAM. And on the package level, it provides a data transmission rate up to 12.8 gigabytes per second (GB/s), which will enable playing of full HD video content in real-time on smartphones and tablets. The high data transfer rates allows LPDDR3 to exceed support for full HD video payback on screens exceeding the current market available four-inch measurement and enable real-time viewing without downloading the content for the full HD visual experience.
By providing an early extensive global supply of LPDDR3 DRAM to smartphone and tablet manufacturers, Samsung expects to see a significant increase in the adoption of 2GB mobile DRAM.
Explore further: NERSC, Cray move forward with next-generation scientific computing