The most logical entry points for enterprise use of SSDs are in storage arrays that combine SSDs with traditional hard drives in a tiered storage environment, in servers that team SSDs with DRAM (dynamic rapid access memory) to extend the rapid processing of dynamic memory, and in stand-alone systems where enterprises simply bring on board racks of SSDs in a distributed storage environment, according to the Storage Networking Industry Association.
"In the next three years, major storage companies will take the concept of SSDs and determine how best to implement it in the enterprise," says Phil Mills, chairman of SNIA's Solid State Storage Initiative. "It is not going to be a case of just replacing hard drives with SSDs, because SSDs will never get down to the cost of hard drives on a dollar-per-GB basis. Instead, the value proposition for SSDs will expand where organizations are trying to realize dramatically improved performance in areas of the enterprise where that improved performance makes sense."
One enterprise area where SSD adoption makes sense is to replace "short stroked" storage. Short-stroking describes a technique in which only the other edges of a series of high-speed Fiber Channel hard disks are formatted, using only 20 percent to 30 percent of the available capacity on each drive. In this way, enterprises are able to improve performance by storing data on the faster part of a drive and by spreading the speed of access across several drives.
Short-stroking is an area where we have seen a lot of enterprise SSD adoption," says Scott Delandy, storage division senior product marketing manager at EMC. "What the enterprise does is take hundreds of Fiber Channel drives out [and] replaces them with four to eight flash drives." Enterprises realize an immediate benefit in total cost of ownership, with lower maintenance and software licensing costs, and also achieve equipment footprint reductions and lower power consumption in data centers.