As a new era of Big Data is being intensified by the rapid adoption of cloud computing, data center density requirements continue to accelerate. Historically, the only way to scale your density was to add more and more slow spinning hard disk drives (HDDs), which has resulted in storage access times that just can’t keep up with the increasing thirst for real-time data analytics. But now there is a more efficient path forward. Extremely high densities are here in the form of solid state drives (SSDs) offering substantial storage improvements in total cost of ownership with high-efficiency designs, major performance gains, significant space savings and unbeatable power utilization.
Hard drive competition
For many years, HDDs have reigned supreme for virtually any enterprise storage need. About a decade ago, SSDs appeared, packing high performance and power efficiency with relatively low latency. The market began to move away from HDDs.
SSDs continued to improve in IOPS processed, as well as the volumes of sequential/random reads and writes they could handle, and how comparatively little power they needed to function. Soon after, 3D NAND manufacturing and triple-level-cell (TLC) technology were introduced adding to the margin of SSD superiority in value and performance. Still many held back, saying the densities were too low from an enterprise-wide TCO perspective. Now that argument is all but erased. SSDs are driving major change in the storage drive space, with densities approaching 16 terabytes in 2.5-inch form factors.
The Density Silver Lining
Currently, SSDs are up to 50 percent larger than HDDs. This is resulting in more than three times the density per storage system in a data center. Yes, triple the capacity! The added spike is due to the fact that the new density champion – our PM1633a SSD with up to 15.36 terabyte (TB) drive – only requires 2.5 inch drive carriers. Why this is important is that most entry level storage systems support either twelve 3.5 inch drive carriers or twenty-four 2.5 inch drive carriers. As you can see, using the smaller drive carriers alone effectively doubles rack storage density.
Let’s show you more of why this has astute CIO and data center managers listening intently.
Because of the significant latencies caused by the slow mechanical process of accessing the data on the disk, IOPS (input/output operations per second) are severely limited when using spinning hard drives. As a result, most storage systems have been deploying more hard drives than are actually required. This is particularly true for deployments of 15K or 10K RPM hard drives. However, to increase system level performance, IT managers often pooled multiple HDDs together (striped) to achieve necessary performance requirements, resulting in lots of wasted storage space and severe under-utilization of data center servers and storage systems.
Due to the remarkable performance improvements and latency reductions offered by SSDs, IT managers are discovering that when you replace HDDs with SSD of the same capacity, very often you can replace more than one HDD with a single SSD. The total number to be replaced would depend upon the end-user workload.
Already Embracing High-density SSDs
Large capacity SSDs and plummeting NAND costs are signifying that the end is near for many HDDs, particularly mission-critical 10K and 15K RPM HDDs.
Several of our customers are already leading the way in deploying high-capacity SAS SSDs in their storage solutions. Are you next?
Tell us, what SSD densities will your next enterprise storage order specify? And how many HDDs can you replace with just one SSD?