Data Backup Digest

Whether you love it or hate it, the big data revolution is here to stay. Industry experts are touting big data as the wave of the future, and even our nation’s top academic institutions are now offering classes in this highly innovative and dynamic field. Not to be outdone, so of the industry’s most prolific manufacturers are starting to base their new hardware designs around the nuances of big data.

Introducing HybriDIMM

You’re probably already familiar with the concept of the DIMM, or dual in-line memory module, platform. After all, the RAM that powers your desktop or laptop computer is probably using the DIMM format. While this type of memory has tremendous advantages over its predecessors, it’s simply not fit to handle the extended datasets of the big data revolution.

Netlist’s answer to the problem comes in the form of their trademarked HybriDIMM storage class memory. Originally unveiled at the 2017 Flash Memory Summit in California, the new chip was designed specifically with advanced artificial intelligence and big data analytics in mind.

C.K. Hong, chief executive officer with Netlist, touted some of the advantages by saying: ''“With the flexibility to operate as main memory, block storage and persistent memory in a standard server – a first in the industry, HybriDIMM can serve as both a memory or storage solution for the rapidly growing markets in big data analytics, in-memory computing and artificial intelligence. This demonstration shows the versatility of HybriDIMM's revolutionary architecture and the practical benefits of data access through multiple operating modes in the memory channel SCM. We are pushing forward with our plans for commercializing this ground-breaking product, in concert with our development partners.”''

Although HybriDIMM is an evolution in its own right, it’s really just the tip of the iceberg when it comes to accommodating big data in the 21st century. With more consumers and enterprises warming up to the idea of big data on a daily basis, and with more companies vying for their money, it will be interesting to watch the progression of big data and the technology that is made to support it.

Technical Specifications

HybriDIMM combines both DRAM and NAND flash memory to bolster the efficiency of exchanging information with the system. It features three different operating modes, including main memory access, block storage and persistent memory, which provide the flexibility needed to handle the various applications of big data seen today.

It also promotes higher access times by accommodating multiple terabytes of persistent memory for real-time applications on Intel servers. The result is a system that is far more efficient than the traditional, NAND-based platform.

The first generation of Netlist’s HybriDIMM features 256 to 512 GB of NAND plus an additional 8 to 16 GB of DRAM. Second generation products feature 1 TB of NAND and 32 GB of DRAM. Both lines of HybriDIMM are currently available on the market.

Find Out More

To find out more information about Netlist, their trademarked HybriDIMM platform or any of their own solutions, please visit their official website at www.netlist.com.