If one trend stands out more in 2006 than any other, it would have to be the way information-technology organizations have been driving the cost out of their hardware to free up more money to invest in applications.
Most of the work thus far has been concentrated around server consolidation efforts; this will soon be augmented by broader adoption of virtualization software, which should significantly improve server utilization rates.
But one hardware area where we have only begun to drive out the cost is storage. It doesnt take a rocket scientist to figure out that there is a wide discrepancy between the kinds of margins that a subsystem manufacturer such as EMC commands versus what a company like Seagate makes by selling the actual disk-drive component.
This is only one of the indicators that suggest there is a lot of room for driving out subsystem cost, regardless of what companies such as EMC, Hewlett-Packard and IBM claim concerning the value of their investments in firmware for their subsystems.
But even more troubling is the fact that while storage space demand is growing 50% to 60% per year thanks to compliance regulations, XML and a rising tide of multimedia files, utilization rates for storage subsystems are still pretty abysmal. For example, utilization rates for server-attached storage are usually about 20% to 30% per array, while rates for network-attached storage (NAS) are roughly 30% to 40% per array. In effect, this means were spending money more inefficiently than ever on storage.
Clearly, these types of utilization rates are a compelling argument for storage virtualization products, which allow multiple arrays to look like a single shared storage resource for multiple servers. But beyond virtualization, companies such as Network Appliance have been making a lot of headway by arguing that customers also need updated storage management software that significantly reduces the number of copies of the files they need to store; the software minimizes overhead as well by copying only the delta changes made to a file rather than making a new copy of the file.