High-volume servers have never been more essential in the Internet Age than now, as exabyte after exabyte of data is being transmitted through data centers week after week. Many factors go into making data move as smoothly as it does. One of the most important is having much faster, more efficient memory that consumes less power. I’m delighted to say that we’re on the threshold of large-scale use of the newest, most important DRAM ever – one which will make servers purr with high-speed satisfaction. I’m talking about DDR4, which holds tremendous promise for both mission-critical and high volume mainstream servers. Samsung began mass producing its DDR4 at the end of August (2013). This ultra-speed memory is being produced in 4Gb chips, using 20 nanometer-class* process technology. It’s the basic building block for a new wave of high-performance, high density DRAM that will spur widespread use of 16GB and 32GB memory modules in servers.
Samsung DDR4 is extremely fast, highly reliable and uses significantly less power than its predecessors. Its current speeds are as high as 2133 megabits per second, aiming towards 2400 mbps later this year. That amounts to transmitting 2.1 billion bits of data every second out of every DDR4 pin. Moreover, DDR4 reduces power consumption by more than 20%, compared to the previous generation of memory. What’s more, Samsung DDR4 comes equipped with some of the best error correction in the business.
Yes, DDR4 is a logical progression in memory development, but because of its superb performance levels, we believe that it will secure a mainstream spot in the enterprise marketplace much faster than many expected.
The rapid adoption of premium 4Gb DDR4 in large-scale data centers and other next-generation server applications will trigger a great deal of interest across the enterprise. Goodbye DDR2 and inefficient versions of DDR3; welcome DDR4. Now, tell us what you think about DDR4 and the role of memory in the data center