Post by Gary Orenstein (thank you)
Flash is a hot topic, but what is the value of flash in the data centre? Is the price such that adoption in the enterprise will continue at a rapid pace? Or is it still a niche offering, only to be used in edge cases where extreme performance is the primary requirement?
The basis for many IT investment decisions is return on investment (ROI). Traditionally, evaluating the benefit falls into two categories: Financial and non-financial.
To encourage prudent use of corporate money, someone, somewhere will have most likely decided that all projects IT or otherwise may only progress if the payback period is “X months” or less.
New photocopiers, telephone systems and replacement SANs – all stand in the same evaluation line. Essentially offering incremental improvements, they each desperately try to shave reductions to the cost per printed page or cost per gigabyte or whatever the applicable metric may be.
But what would the dynamics of a data centre look like with the introduction of flash-based storage memory? Holding the “hot” or process-critical data on flash (with disk taking a more comfortable role as “the new tape” holding the cold data) yields results that upset the status quo and create the conditions for significant improvements in low-latency data performance while offering compelling ROI. This is not only in terms of work done, but also for those who wish to reduce capital and operational expenses related to real estate, energy usage and software licensing.
Read on here