Justifying the Cost of SSDs
Gregory Wong with analyst group Forward Insights says that “SSDs use non-volatile NAND flash memory chips, which are cheaper than DRAM chips but are still as much as 18 times more expensive than 15,000 rpm Fibre Channel or serial SCSI (SAS) drives.” He claims that the price of SSDs is expected to come down as more customers adopt solid state drive and NAND flash card technology, a fact that will trickle down to the data center.
The breaking point for SSD affordability will be the magical point in time when price reaches the $1 per gigabyte level, which some analysts expect will be late next year. As it stands now, the price for NAND flash in an SSD form factor is around $9 per gigabyte for a high-end, single-level cell flash and $3 per gigabyte for multi-level cell flash. To put this in perspective, Fibre Channel or SAS drives are in the 50 to 60 cents per gigabyte range.
As Lucan Mearian said, “When it comes to PCIe NAND flash cards, like those sold by Fusion-io, Texas Memory Systems, Micron or Virident Systems, which can be used in all-flash arrays or in application servers themselves, prices can go through the roof, but so does performance due to the higher speed interconnect and the proximity of the flash storage to the server processors.”
The problem is, some people don’t understand why the expense is there and furthermore, might not understand that this is not a magic bullet for dramatic performance improvements. In fact, some use cases offer examples where the investment in SSD technology is far from worth it.
From an example of this point in ComputerWorld
Dan Marbes, a systems engineer at the Green Bay, Wis.-based bank, decided to try solid-state drives (SSD) to increase the performance on I/O-hungry applications, while reducing his spindle footprint.
He bought three SSDs to serve as top-tier storage for business intelligence (BI) applications on his SAN. The flash storage outperformed 60 15,000-rpm Fibre Channel disk drives when it came to small-block reads.
However, when Marbes used the SSDs for large-block random reads and any writes, “the 60 15K spindles crushed the SSDs,” he said, demonstrating that flash deployments should be strategically targeted at specific applications.
The idea here is that SSDs can easily cover their own costs, sometimes in less than a year, depending on use. However, since this is not a perfect one-size fits all solution for all, those who can’t clearly benefit from SSDs are the ones most likely to express concern about what seems like a ridiculous cost.
Full story at ComputerWorld